Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Recent Posts
1
3rd Party Plugins - (fb2k) / Re: foo_skip: skip tracks that match a specified search query
Last post by padampadam -
I use this plugin on a daily basis and I love it. Here's a big thank you to all who contributed!

If I could suggest something it would be to have a toggle on/off option, or even better, a toolbar with a dropdown menu containing registered skip queries that we could switch on the fly, or even disable. And also an option to clear past entries, it can get confusing really fast.
2
3rd Party Plugins - (fb2k) / Re: ReplayGain DSP - Alternative ReplayGain implementation by Case
Last post by Defender -
Quote from: gorman on
Hi! Thanks for the work on this plugin, I especially like the approach similar to Lyrion Music Server "Smart replaygain".

I have a (hopefully) small feature request: would it be possible to have an option to do nothing if a file lacks replaygain info?

In my case I listen to DTS and AC3 files that lack that and, to bitstream them, no processing has to happen in foobar2000.
How do you bitstream DTS and AC3 from within foobar?

I cannot and therefore I just added replaygain values to DTS and AC3 files.
3
Lossless / Other Codecs / Re: New lossless audio codec in development
Last post by genuine -
Quote from: Hakan Abbas on
DLP is completely different from Levinson-Durbin. However, it can be thought of as a dynamic version of fixed estimators. The learning mechanism in DLP is actually trying to find the best case for a block selected from zero (e.g. 512 samples), i.e. the case with the least error. Here a decision is made by looking at only 2 or 3 samples. And the mistakes made are tried to be improved. This is not very interesting.

But what is interesting is that once the appropriate parameters are set, the best result can be obtained with the same parameters in the previous or subsequent blocks. Even if this depends on the shape of the data, in my tests it can sometimes be valid for hundreds of blocks before or after. you can see this by trying it immediately. Maybe we will also see the negative aspects of the method.
In my experiments I used blocks of 500 samples. In DLP, if the best parameters are determined for each block, the error averages can be much lower. However, even with the same parameters, it seems that many consecutive blocks can operate with similar efficiency. If there are no inaccuracies, I have also found that fixed estimators give very good results for the music I have selected according to Levinson. And they are inexpensive. But Levinson gives better results than fixed estimators and close to DLP on larger blocks.

https://www.rarewares.org/test_samples
ATrain.wav
DLP 1.9 / 0.9 and Levinsion Degree 10

Quote
Range 120,000 - 120,500
P-1: (Absolute Error Avarage) |E|= 443.106   (Root Mean Square Error) RMSE= 562.444
P-2: (Absolute Error Avarage) |E|= 207.323   (Root Mean Square Error) RMSE= 278.034   
P-3: (Absolute Error Avarage) |E|= 233.894   (Root Mean Square Error) RMSE= 303.262   
P-4: (Absolute Error Avarage) |E|= 337.748   (Root Mean Square Error) RMSE= 428.854   
L-D: (Absolute Error Avarage) |E|= 385.817   (Root Mean Square Error) RMSE= 512.501   
DLP: (Absolute Error Avarage) |E|= 149.771   (Root Mean Square Error) RMSE= 196.561   

Range 123,000 - 123,500
P-1: (Absolute Error Avarage) |E|= 310.696   (Root Mean Square Error) RMSE= 405.519   
P-2: (Absolute Error Avarage) |E|= 125.683   (Root Mean Square Error) RMSE= 167.74   
P-3: (Absolute Error Avarage) |E|= 121.777   (Root Mean Square Error) RMSE= 153.605   
P-4: (Absolute Error Avarage) |E|= 163.332   (Root Mean Square Error) RMSE= 200.383   
L-D: (Absolute Error Avarage) |E|= 301.466   (Root Mean Square Error) RMSE= 393.957   
DLP: (Absolute Error Avarage) |E|= 81.6198   (Root Mean Square Error) RMSE= 104.565   

Range 130,000 - 130,500
P-1: (Absolute Error Avarage) |E|= 257.808   (Root Mean Square Error) RMSE= 346.281   
P-2: (Absolute Error Avarage) |E|= 134.14   (Root Mean Square Error) RMSE= 185.864   
P-3: (Absolute Error Avarage) |E|= 153.179   (Root Mean Square Error) RMSE= 197.29   
P-4: (Absolute Error Avarage) |E|= 225.078   (Root Mean Square Error) RMSE= 281.94   
L-D: (Absolute Error Avarage) |E|= 260.342   (Root Mean Square Error) RMSE= 345.381   
DLP: (Absolute Error Avarage) |E|= 106.201   (Root Mean Square Error) RMSE= 134.686   

Range 150,000 - 150,500
P-1: (Absolute Error Avarage) |E|= 191.308   (Root Mean Square Error) RMSE= 247.628   
P-2: (Absolute Error Avarage) |E|= 128.058   (Root Mean Square Error) RMSE= 159.02   
P-3: (Absolute Error Avarage) |E|= 175.956   (Root Mean Square Error) RMSE= 222.19   
P-4: (Absolute Error Avarage) |E|= 270.7   (Root Mean Square Error) RMSE= 338.797   
L-D: (Absolute Error Avarage) |E|= 196.845   (Root Mean Square Error) RMSE= 253.754   
DLP: (Absolute Error Avarage) |E|= 96.0289   (Root Mean Square Error) RMSE= 121.365   

Range 200,000 - 200,500
P-1: (Absolute Error Avarage) |E|= 174.672   (Root Mean Square Error) RMSE= 214.922   
P-2: (Absolute Error Avarage) |E|= 171.539   (Root Mean Square Error) RMSE= 213.147   
P-3: (Absolute Error Avarage) |E|= 257.679   (Root Mean Square Error) RMSE= 320.909   
P-4: (Absolute Error Avarage) |E|= 407.173   (Root Mean Square Error) RMSE= 510.26   
L-D: (Absolute Error Avarage) |E|= 179.434   (Root Mean Square Error) RMSE= 222.833   
DLP: (Absolute Error Avarage) |E|= 145.787   (Root Mean Square Error) RMSE= 181.068   

And Range 200,000 - 200,500 for DLP 1.2 / 0.9
DLP: (Absolute Error Avarage) |E|= 119.179   (Root Mean Square Error) RMSE= 146.419   

https://www.rarewares.org/test_samples
Bachpsichord.wav
DLP 0.9 / 0.6 and Levinsion Degree 10

Quote
Range 781,000 - 781,500
P-1: (Absolute Error Avarage) |E|= 323.39   (Root Mean Square Error) RMSE= 401.816   
P-2: (Absolute Error Avarage) |E|= 246.164   (Root Mean Square Error) RMSE= 308.946   
P-3: (Absolute Error Avarage) |E|= 264.817   (Root Mean Square Error) RMSE= 332.567   
P-4: (Absolute Error Avarage) |E|= 319.684   (Root Mean Square Error) RMSE= 406.375   
L-D: (Absolute Error Avarage) |E|= 328.579   (Root Mean Square Error) RMSE= 401.791   
DLP: (Absolute Error Avarage) |E|= 236.809   (Root Mean Square Error) RMSE= 295.039   

Range 783,500 - 784,000
P-1: (Absolute Error Avarage) |E|= 576.748   (Root Mean Square Error) RMSE= 732.896   
P-2: (Absolute Error Avarage) |E|= 739.136   (Root Mean Square Error) RMSE= 923.171   
P-3: (Absolute Error Avarage) |E|= 1188.61   (Root Mean Square Error) RMSE= 1482.96   
P-4: (Absolute Error Avarage) |E|= 2095.01   (Root Mean Square Error) RMSE= 2596.91   
L-D: (Absolute Error Avarage) |E|= 565.347   (Root Mean Square Error) RMSE= 722.512   
DLP: (Absolute Error Avarage) |E|= 599.783   (Root Mean Square Error) RMSE= 758.233   

Range 785,000 - 785,500
P-1: (Absolute Error Avarage) |E|= 858.63   (Root Mean Square Error) RMSE= 1076.89   
P-2: (Absolute Error Avarage) |E|= 1068.71   (Root Mean Square Error) RMSE= 1312.39   
P-3: (Absolute Error Avarage) |E|= 1539.71   (Root Mean Square Error) RMSE= 1930.56   
P-4: (Absolute Error Avarage) |E|= 2525.31   (Root Mean Square Error) RMSE= 3107.55   
L-D: (Absolute Error Avarage) |E|= 801.279   (Root Mean Square Error) RMSE= 1003.89   
DLP: (Absolute Error Avarage) |E|= 774.577   (Root Mean Square Error) RMSE= 961.112   

Range 789,000 - 789,500
P-1: (Absolute Error Avarage) |E|= 541.868   (Root Mean Square Error) RMSE= 674.853   
P-2: (Absolute Error Avarage) |E|= 560.549   (Root Mean Square Error) RMSE= 700.241   
P-3: (Absolute Error Avarage) |E|= 763.349   (Root Mean Square Error) RMSE= 948.568   
P-4: (Absolute Error Avarage) |E|= 1119.4   (Root Mean Square Error) RMSE= 1435.04   
L-D: (Absolute Error Avarage) |E|= 533.588   (Root Mean Square Error) RMSE= 670.675   
DLP: (Absolute Error Avarage) |E|= 439.466   (Root Mean Square Error) RMSE= 548.961   

Range 795,500 - 796,000
P-1: (Absolute Error Avarage) |E|= 305.056   (Root Mean Square Error) RMSE= 387.991   
P-2: (Absolute Error Avarage) |E|= 269.337   (Root Mean Square Error) RMSE= 337.022   
P-3: (Absolute Error Avarage) |E|= 321.195   (Root Mean Square Error) RMSE= 393.435   
P-4: (Absolute Error Avarage) |E|= 412.865   (Root Mean Square Error) RMSE= 506.12   
L-D: (Absolute Error Avarage) |E|= 306.578   (Root Mean Square Error) RMSE= 390.907   
DLP: (Absolute Error Avarage) |E|= 229.701   (Root Mean Square Error) RMSE= 290.912   

And Range 795,500 - 796,000 for DLP 1.5 / 0.8
DLP: (Absolute Error Avarage) |E|= 174.152   (Root Mean Square Error) RMSE= 207.577   
4
3rd Party Plugins - (fb2k) / Re: ReplayGain DSP - Alternative ReplayGain implementation by Case
Last post by gorman -
Hi! Thanks for the work on this plugin, I especially like the approach similar to Lyrion Music Server "Smart replaygain".

I have a (hopefully) small feature request: would it be possible to have an option to do nothing if a file lacks replaygain info?

In my case I listen to DTS and AC3 files that lack that and, to bitstream them, no processing has to happen in foobar2000.
5
General - (fb2k) / Re: foobar2000 minimalist layout
Last post by anamorphic -
You want a library viewer that can display covers. Maybe something like Coverflow (this one is a mod for 32/64-bit that can also follow playlists, discussion topic here).

I'm not sure what your remote control buttons are mapped to for scrolling to the next album? - Coverflow can use left/right arrow keys for that, besides mouse wheel.

The preferences for it are a little confusing. It looks like if you right-click the Coverflow window > Library Selections > Set Library Selection, it appears to update the library playlist to that album when you scroll, then hitting 'Play' should play it. (You might need Playback menu > Playback Follows Cursor on).
6
General - (fb2k) / foobar2000 minimalist layout
Last post by il_fait_beau -
Hello,

I'm new here and i thank you for all the informations i can find.

My hifi system is laptop, dac, preamp, amp, and speaker, and the use of the preamp remote to control the listening(foobar and preamp).
I use foobar2000 in latest version 2..24.3 x64 on w11 for its black theme.
The remote is setup using flirc usbA, superposing laptop keyboard.

I can only use 7 remote buttons to control foobar: track<, track >,search<<,search>>, play, pause and stop.

At the moment my minimalist layout needs 2 UI element , album list(sort by folder structure) and album viewer.
So i can scroll the albums, see their pictures, play, pause stop and have next&previous track.

What i wish would be the possibility to have the same fonctions without the need of UI element album list.

Why ? because i sit 4 meters from the speaker and i cannot read the laptop screen, i only need to see  the front album Art
and scroll up and down.

How would you setup foobar for that ?


 














7
Lossless / Other Codecs / Re: New lossless audio codec in development
Last post by SebastianL -
Next step in lossless audio compression should involve
1. merging prediction + residual encoding -> model p(x_t | x_{t-1}, x_{t-2}, ...} directly
2. use lstm/gru for prediction, e.g. wavenet: https://arxiv.org/abs/1609.03499

Issue with 2. is that for a frame the NN does not converge fast enough using (stochastic) gradient descent.
Possible solutions:
- use a hybrid approach, where you load a (large) base model from disk and update during prediction
- train a (small) NN on every frame and save it -> have to find a compromise between model-size, accuracy and final file size

Having a large base model opens a discussion how to compare such variants to classical (statistical) codecs
See https://en.wikipedia.org/wiki/Kolmogorov_complexity