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 486515 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

Re: Resampler plugin

Reply #125 –
I maybe being an idiot - it has been known - but I can't configure the SoX component in the 'Convert...' dialogue box in f2k v1.1
I can use the component in real time and I must say I'm really impressed with the quality from 24b96k to 16b44k... really impressed.

I'm running f2k v1.1 on win7 Pro (x64), could this be a reason I can't configure the component.


Cheers

Re: Resampler plugin

Reply #126 –
What happens when you press "Configure selected" button?

Re: Resampler plugin

Reply #127 –
Quote from: lvqcl on
What happens when you press "Configure selected" button?

Guess what, It was me being an idiot. A little further digging around led to me realising I need to use the 'DSP Chain Preset'.
I successfully converted a file from 96KHz to 44.1KHz. The bit depth didn't change however from 24bit to 16bit. Is that because I used 'Very High' setting?

BTW... the 'configured selected' button is always greyed out.

Re: Resampler plugin

Reply #128 –
Quote
The bit depth didn't change however from 24bit to 16bit. Is that because I used 'Very High' setting?

No. Click on Output format link in main converter window and set output bit depth to 16.


Quote
BTW... the 'configured selected' button is always greyed out.

Select "SoX Resampler" entry in Active DSPs list and this button should be enabled.

Re: Resampler plugin

Reply #129 –
Sorry, it still stays greyed out. I've used the DSP chain option and it worked like a treat that way though.


Great stuff, thanks for getting SoX onto windows.

Re: Resampler plugin

Reply #130 –
Thank you for your resampler, it has increased the quality but it also adds a small "click" at the beginning.

Configuration : Foobar 1.1.1 with resampler 0.5.5 mod (target resample rate: 176400)
Pc with Intel i5 , USB output in KS with Hiface and V-DAC

Can you do something ?

Thanks in advance

Re: Resampler plugin

Reply #131 –
Cannot reproduce the issue. Does the 'click' exist with PPHS or other resamplers?

Re: Resampler plugin

Reply #132 –
Quote from: lvqcl on
Cannot reproduce the issue. Does the 'click' exist with PPHS or other resamplers?

Yes, but I also noticed recently that the same exists with any DSP plugin. This is probably due to the way Foobar manages the plugins in my particular configuration with the Hiface in Kernel Streaming but I am not expert enough to draw any conclusion. Could be also Hiface's fault.

Have you tried also with a Hiface ?

Re: Resampler plugin

Reply #133 –
Not everyone has a Hiface, try switching back to DirectSound to see if you still have that particular problem. Also are you using the new VST bridge?

Re: Resampler plugin

Reply #134 –
::

Which one should I use (0.5.4.4 or 0.5.5) for INTEL SSE2 Pentium 4 non-HyperThreading processor?


Greetings ...

::

Re: Resampler plugin

Reply #135 –
0.5.5 requires SSE3 so you don't have choice

Re: Resampler plugin

Reply #136 –
::

So the bold should be "requires".

Thanks a lot!   

::

Re: Resampler plugin

Reply #137 –
So the new foobar2000 resampler should work fine on AMD Athlon PCs ?

Does it clash with any version of Windows?

Re: Resampler plugin

Reply #138 –
0.5.4.4 works on any CPU, 0.5.5 requires SSE3 (so it works on almost all modern CPUs).

And of course it doesn't have any particular OS requirements.

Re: Resampler plugin

Reply #139 –
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.

Re: Resampler plugin

Reply #140 –
Wow, what a thread! 

Quote
PROS:
-Slightly smoother treble, and definitely less grainy
-Clearer low-level details and easier to hear ambience of the recording
-Larger soundstage (because of the above point)
-Blacker background
-Better dynamics

CONS:
-Slightly less bass

Quote
the better focus, clearer vocals, deeper soundstage, better imaging, ambience details retrieval and a sense of "faster" sound and musical flow. but i noticed the sound significantly lost on the weight, cello and violins immediately struck me as too thin sounding and unnatural.


Well, for 44.1 => 96kHz upsampling  the frequency response of SoX resampler is almost flat for 0...20.5 kHz (default settings). So it doesn't attenuate low frequencies.


BTW, version 0.5.6 is out.

Re: Resampler plugin

Reply #141 –
Thanks for the swift reply! Yes, I know that sounds weird, but I'd dare saying that the deep bass appears to be less percussive. I've tried it on some old school hip hop, and the TR808 kick drums seem less percussive indeed. But that was with 0.55, I'll try again tomorrow with fresh ears using 0.56.

Re: Resampler plugin

Reply #142 –
Fresh ears may or may not help. Unless you are doing a blind test you may hear differences that do not exist. If you expect to hear a difference your brain can fabricate it. You must perform a blind test to avoid that. This has been nicely demonstrated in Ethan Winer's "Audio Myths Workshop" video from the 2009 AES conference: http://www.youtube.com/watch?v=BYTlN6wjcvQ

The audio device may produce a different frequency response when its HW is set to a different sample rate, but usually these differences are not big enough to be audible. However, to completely eliminate the possible effect of the HW working differently with different sample rates you can prepare and compare audio samples that have the same final sample rate.

For instance, if you are resampling from 44.1 to 96 KHz, a good test would be to resample a 44.1 KHz audio sample to 96 KHz once and compare it with a sample that has been resampled three times: 44.1 -> 96 -> 44.1 -> 96 KHz. If resampling causes an audible difference, the difference should more pronounced in the latter sample. You can include a resampler in the DSP options when you convert files with foobar.

Create the two resampled audio files. Compare them with foobar's ABX comparator. If you can detect a difference in a blind ABX test post the test log here and a short clip of the original unresampled audio file (max. 30 s) to the uploads forum.

Re: Resampler plugin

Reply #143 –
yeah, thanks for the instructions!

but tbh, the THD in the trebles is too high, and the low end bass sounds less percussive and dull. It still sounds much better than PPHS, this one sounds horrid.

I fully agree with the guys on head-fi, it's very nice for 1h then it gets odd. Upsampling doesn't come for free, never did, never will.

Re: Resampler plugin

Reply #144 –
What is the difference of MOD2 and MOD1?

PPHS(ultra)
Speed (x realtime): 26.301

SSRC X
Speed (x realtime): 30.351

MOD 1
Speed (x realtime): 67.376

MOD 2
Speed (x realtime): 366.118






Re: Resampler plugin

Reply #147 –
Quote from: bison on
yeah, thanks for the instructions!

but tbh, the THD in the trebles is too high, and the low end bass sounds less percussive and dull. It still sounds much better than PPHS, this one sounds horrid.

I fully agree with the guys on head-fi, it's very nice for 1h then it gets odd. Upsampling doesn't come for free, never did, never will.

Please watch the video in my link and try the ABX test.

In my experience the mentioned resamplers do not audibly alter the audio quality when they resample from one sample rate to a higher sample rate. A while ago I measured some upsampled test files. With the default settings SoX produced a lot less ringing and other artifacts than PPHS at about 20 kHz, but even the distortion that PPHS produced was likely to be completely inaudible because of the very high frequency. PPHS has been designed to be fine for playback purposes. If you can prove otherwise, please post your ABX logs and test samples.

I just did a quick test. I compared two "Sox" samples. The first one was upsampled once from 44.1 to 96 kHz (using the 24-bit output mode to avoid any increased noise floor) and the second sample was upsampled four times: 44.1 -> 48 -> 64 -> 88.2 -> 96 kHz (all 24-bit).

I couldn't hear any difference in an ABX test. I also measured the difference signal of these two files with Adobe Audition. Its maximum amplitude was -129 dB below 0 dBFS at about the 19.5 - 21.5 kHz frequency range. I.e. much below anything that can be audible. The lower frequencies were even quieter. I can upload the samples if anyone wants to try them.

In general, upsampling is useful only in certain cases when a particular hardware device does not perform well (or not at all) at a certain sample rate.

(Of course you may have other good reasons to resample than audio quality. You may have a receiver that supports only certain sample rates, you may want to mix various audio streams that are of different sample rates, etc.)

Re: Resampler plugin

Reply #148 –
Version 0.6.0 was released.

Re: Resampler plugin

Reply #149 –
Hi lvcql,

Thanks for all your work on this plugin. 

My media PC is connected to a Meridian audio processor and I currently use your plugin to upsample all my lossless CD rips to 88.2kHz in the hope that minimizes the effects of any jitter in the SPDIF connection between my soundcard and the processor.

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.

Is there any possibility you could extend the minimum allowed passband in your plugin down to 85% so that I could try some of the low passband combinations mentioned in the link?

Thanks.

-Ronnie