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.
Topic: Resampler plugin (Read 490083 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: Resampler plugin

Reply #150
Thank you for this great sounding plugin!

In this thread, one slight drawback was pointed out, it's that it seems to attenuate the bass response a bit: http://www.head-fi.org/forum/thread/467853...r-sox-resampler

Any thoughts on that matter please? Because I tend to agree.

When I switched to Sox, I noticed the same. Somehow a slightly weaker bass and mids were more prominent.
It's the Phase response. It has to do with pre-echo/post-echo, that changes the sound of the audio. (Sorry, my english is to bad to explain this in detail.)
With Phase response to 25% (intermediate), I've got my usual sound back. Nice bass and highs. You may try some Phase response values between 10% - 40%. Maybe it sounds better on your system.

Sox manual:
A phase response setting may be used to control the distribution
of  any  transient  echo  between ‘pre’ and ‘post’: with minimum
phase, there is no pre-echo but the longest post-echo; with lin-
ear  phase,  pre  and  post echo are in equal amounts (in signal
terms, but not audibility terms); the intermediate phase setting
attempts to find the best compromise by selecting a small length
(and level) of pre-echo and a medium lengthed post-echo.



Re: Resampler plugin

Reply #153
Is there a reason to update to version 0.6.0? Changelog? Version 0.5.4.4 works perfect on my System.

Re: Resampler plugin

Reply #154
Changelog is in the post #2 in the upload thread.

0.5.6 (and 0.6.0) are noticeably faster, at least on Intel Core 2 and Core i.


Re: Resampler plugin

Reply #156
From sox.pdf:
Quote
The −b option allows the band-width to be set to any value in the range
74−99.7 %, but note that band-width values greater than 99% are not recommended for normal use
as they can cause excessive transient echo

Re: Resampler plugin

Reply #157
Thanks for explanation. So you can add that 0.2% yet..

Re: Resampler plugin

Reply #158
Meridian and some other high-end companies like Ayre claim that their new upsampling "apodizing" filters can help sound quality by minimizing pre-echo.  Unfortunately, my older Meridian processor doesn't implement these new upsampling filters.

Someone on the Meridian forums has been trying to reproduce a filter which has an impulse response like the Meridian filter using a SOX plugin for their squeezebox server, see: http://www.meridianunplugged.com/ubbthread...5273#Post125273

Unfortunately, the SOX filter which has an impulse response which looks closest to the Meridian one has a passband parameter of 87.5% which is lower than the GUI for your plugin allows.  From reading the thread on the Meridian forums it looks like setting a low passband on these filters helps shape how fast the post-echo tails off.


All over the net i see that funny graphs of pre and post echo. Can anyone help me interpreting these? As i interprete these graphs the post and pre echo that these pics show consist of frequencies "above the cutoff" that is applied. So they are out of the audible range. I wonder why people hear all kinds of problems with noise that happens above there at a very low amplitude. To get rid of that people use low-pass filters that come in way deeper in frequency and therefore may alter the sound much more as the original echo. Am i wrong? Of cause from a marketing point pre-echo is evil and can be easily shown with a simple graph. I wonder how many % of the people looking at these pics realize the shown echo is happening above their hearing.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Re: Resampler plugin

Reply #159
Is is right that the normal version does resampling even if input = output samplerate? I need to use the "mod" version, correct? Shouldn't this get "fixed" in SoX then?

Btw: What's the bitdepth of all the DSPs in foobar2000? 64bit float?

Re: Resampler plugin

Reply #160
Is is right that the normal version does resampling even if input = output samplerate? I need to use the "mod" version, correct?

Of course, it doesn't. So you don't need mod version.

Btw: What's the bitdepth of all the DSPs in foobar2000? 64bit float?

Input and output data are 32-bit float.


Re: Resampler plugin

Reply #162
Where does the mod version store its configuration data?

The normal version stores it in "foo_dsp_resampler.dll.cfg". But I can't find such a file for the Mod version!?



Why does the Mod version gives information about resampling in the Console but the normal version doesn't?

Re: Resampler plugin

Reply #163
Ok, thanks! Could swear I've read about 64bit float processing somewhere ...

That was true for foobar2000 0.8.x (and maybe older versions too)


Where does the mod version store its configuration data?

The normal version stores it in "foo_dsp_resampler.dll.cfg". But I can't find such a file for the Mod version!?

DSP settings are stored in DSP presets (in Core.cfg, I assume). 'Normal' version has also some player-wide settings and they are stored in foo_dsp_resampler.dll.cfg file.


Why does the Mod version gives information about resampling in the Console but the normal version doesn't?

Why not? 

Re: Resampler plugin

Reply #164
Ok, thanks! Could swear I've read about 64bit float processing somewhere ...

That was true for foobar2000 0.8.x (and maybe older versions too)

Why the downgrade?

Where does the mod version store its configuration data?

The normal version stores it in "foo_dsp_resampler.dll.cfg". But I can't find such a file for the Mod version!?

DSP settings are stored in DSP presets (in Core.cfg, I assume). 'Normal' version has also some player-wide settings and they are stored in foo_dsp_resampler.dll.cfg file.

I see.

Why does the Mod version gives information about resampling in the Console but the normal version doesn't?

Why not? 

Ok. Might make sense for those mod versions.


Re: Resampler plugin

Reply #166
Weird (as most DAWs nowadays also went to 64bit float processing).

Re: Resampler plugin

Reply #167
lvqcl, i've some comments on you. I've discovered, you created a clone of srcDrop (rateDrop), which uses sox resampling instead of libsamplerate.
Thank you for this tool, i mean it was really great idea to pair simplicity gui with sox resampling. I haven't  found a special thread about it, therefore i ask here.

1. if there is also bitwidth change applied, does it use any kind of dithering/noise shaping? it doesn't seem so.
2. on this site http://leute.server.de/wilde/resample.html there stays, sox used polyphase filtering as one of the three resamp. algorithms. Its quite old article, so it seems polyphase filtering was later excluded. If its true, why?
3. in the case of sox, does the -s option have any other function than only extend default band-width to 99%? I mean steepness of the filter and band-width could be two different things.
4. does sox rate use 64 or 32-bit float processing? (know fb2k has 32 float)
5. is rateDrop also cpu optimized as fb2k plugin?
I'm sorry if there is discussion already, that answers some questions, haven't found any, though there are some basics in rate.txt article. 

Wishlist for the next version(s):
to extend rateDrop settings to full sox rate effect possibilities, i mean mainly:
-b 74-99.7             
-p 0-100
adding dithering options from sox

Re: Resampler plugin

Reply #168
Honestly, rateDrop is just a byproduct of my development of foo_dsp_resampler. I don't have plans to develop it further.
a) SoX resampling algorithms use 64-bit floating point arithmetics.
b) Latest version of rateDrop uses SSE3 optimisation, if possible.

About SoX itself: I am not an author of it so my knowledge is limited.
a) Yes, -s is the same as -b 99.
b) From SoX changelog: "Improved `rate' resampling effect; resample, polyphase, & rabbit now deprecated." So it was removed in favor of 'rate' effect.
c) SoX converts its input to 32-bit integers anyway.

Re: Resampler plugin

Reply #169
Quote
I don't have plans to develop it further.

It's a pity.
Thanks for your comment.

Re: Resampler plugin

Reply #170
Why not calling the DLL foo_sox_resampler.dll instead of foo_dsp_resampler.dll ?

Re: Resampler plugin

Reply #171
i have tried to download and use version 0.6.0 and 0.5.6
both versions gave foobar an error that said 'the dll is either the wrong format or corrupted'

does it have anything to do with the fact that the file is not the right format when downloaded?
the 0.6.0 version said it was a '0' file
the 0.5.6 version said it was a '6' file

i tried to add .dll to the end of the filename, but apparently it didnt work.

i have used a calibrated microphone to adjust my equalizer.. and then i used DRC to use the digital room correction to further flatten the frequency response.
but the DRC impulse filter is for 48khz
so i am trying to resample everything to 48khz to make things match.

Re: Resampler plugin

Reply #172
Try to add .zip extension to downloaded files and unpack them.

Re: Resampler plugin

Reply #173
I need information to do sample rate conversion. 44.1 kHz -> 48 kHz.

I know that there is foobar plugin but I need it to be done with command line.
I download SOX from here http://sourceforge.net/projects/sox/files/sox/14.3.1/

Highest quality is needed no matter what speed is.
I'm not sure if these are right parameters
Code: [Select]
sox Castanets.wav -b 16 resampled.wav rate -v 48000 dither -s


1.  Is dither must be here in (44.1 kHz ,16 bits) -> (48 kHz ,16 bits) conversion?
2.  What does  -b 16 do exaclty?
3.  -v stands for VHQ (Very High Quality) but do I use it correctly in mentioned command line?

Re: Resampler plugin

Reply #174
1. Probably yes. Note that -s also means noise-shaping: quantization noise will be shifted to high frequencies (>16 kHz)
2. Number of bits in the output file. Redundant, if castanets.wav is also 16-bit.
3. Yes.