Topic: How does JRiver up sample audio...? (Read 17505 times)

Wungun · « **on:** March 09, 2014, 06:29:15 am »

I've heard it uses SRC alogorithm?

Hendrik · « **Reply #1 on:** March 09, 2014, 06:47:29 am »

Resampling is based on Shibatch's SSRC resampler, a well respected implementation in the industry.

Wungun · « **Reply #2 on:** March 09, 2014, 06:51:21 am »

Awesome! Thanks for the reply...

6233638 · « **Reply #3 on:** March 09, 2014, 07:43:00 am »

Do you think alternative resamplers might be a possibility at some point in the future?
Looking at some comparisons (though this is downsampling rather than upsampling) it seems that SoX's VHQ resampler options measure better than SSRC.
SoX also offers intermediate and minimum phase filters, which some people may prefer - especially when upsampling to DSD.

I don't know if their license would permit for its inclusion in a product like Media Center though.

Wungun · « **Reply #4 on:** March 09, 2014, 08:02:28 am »

If you can find it in a compatible VST plugin form, it can be used in JRiver.
Other upsamplers out there too...
I'm equally curious.

Hendrik · « **Reply #5 on:** March 09, 2014, 08:10:48 am »

This is what Matt said to the same question (optional SoX alternative) just in November:

Quote from: Matt

Our resampler is based on Shibatch's SSRC. It's well respected and instrument tests really well.

I'm not eager to chase other options unless there's some clear technical benefit.

If you're interested, you can apply resampling during conversion. This means you can feed tones, sweeps, etc. through it and analyze the output.

6233638 · « **Reply #6 on:** March 09, 2014, 11:54:22 am »

Well I am by no means an expert in this area.
One of the big claims in favor of DSD, is its ability to represent an impulse without any pre or post ringing.

So it seemed that using a minimum phase resampler with no pre-ringing might be advantageous when upsampling to DSD.

However, I have just tried out another product which lets you choose from all kinds of dithering/noise shaping algorithms, upsampling filters (linear phase, minimum phase, apodizing etc.) and I'm not really hearing a reason to change what Media Center is already doing.

mark_h · « **Reply #7 on:** March 25, 2014, 04:27:36 am »

Apodizing is very much something I'd like to see in Media Center. The shortcomings of 44.1/16 sampling are well understood, in particular the unnatural pre-ringing caused by the passband filter. Apodizing during upsampling is a way to "fix" this and that a company like Meridian has gone to the major effort of introducing it into their kit suggests that it has some value to add...

Pono are also adding such filters to their upcoming media player.

AndyU · « **Reply #8 on:** March 25, 2014, 08:08:17 am »

Quote from: mark_h on March 25, 2014, 04:27:36 am

Apodizing is very much something I'd like to see in Media Center. The shortcomings of 44.1/16 sampling are well understood, in particular the unnatural pre-ringing caused by the passband filter. Apodizing during upsampling is a way to "fix" this and that a company like Meridian has gone to the major effort of introducing it into their kit suggests that it has some value to add...

Pono are also adding such filters to their upcoming media player.

How would an apodizing filter in MC interact with whatever filter is already in the DAC?

mark_h · « **Reply #9 on:** March 25, 2014, 08:33:23 am »

Apodizing is only applied during upsampling, so an Apodizing DAC would do no addition apodizing with an upsampled signal. And presumably filters could be defeatable anyway as part of MC's DSP engine should there be any issues.

Wungun · « **Reply #10 on:** March 30, 2014, 07:26:17 pm »

Is there a VST plugin that does this maybe...?

theoctavist · « **Reply #11 on:** March 30, 2014, 08:47:04 pm »

Quote from: Wungun on March 30, 2014, 07:26:17 pm

Is there a VST plugin that does this maybe...?

there is a setting in the SOX resampler for apodizing.

lc flc * *
# FT:{START=--skip=%t}U:{END=--until=%v}
[flac] -dc --totally-silent $START$ $END$ -- $FILE$ | [sox] -q --replay-gain album -t wav - -t flac -C 0 -b 24 - gain -1 rate -v -M -b 87.5 88200 dither -S

39892 · « **Reply #12 on:** April 01, 2014, 02:54:33 am »

Quote from: Wungun on March 30, 2014, 07:26:17 pm

Is there a VST plugin that does this maybe...?

It's my understanding that the VST spec doesn't allow plugins to change the sample rate. That's why you don't find an Izotope SRC plugin for DAWs.

Mikkel · « **Reply #13 on:** April 01, 2014, 04:45:21 am »

Comparing SSRC High Precision and Sox 14.4 Very High Quality: http://src.infinitewave.ca/

Lower distortion and noise floor seems to be the only benefits with SOX. However, they are both at or below -130 db so I am not really sure what is gained in practice?
In the other tests at the site SSRC is either as good as or better than Sox.

Regarding minimum or linear phase... I have never been able to hear the difference my self but I'm open for the suggestion that it can possibly make a difference. In building FIR filters for my setup I have compared minimum and linear phase filters and in this particular setup I cannot hear a difference at all.

Best regards,
Mikkel

theoctavist · « **Reply #14 on:** April 03, 2014, 02:25:50 pm »

Quote from: 39892 on April 01, 2014, 02:54:33 am

It's my understanding that the VST spec doesn't allow plugins to change the sample rate. That's why you don't find an Izotope SRC plugin for DAWs.

there are sr plugins i think. r8 brain and r8 brain pro? or are they standalone only.

6233638 · « **Reply #15 on:** April 04, 2014, 12:10:17 am »

Quote from: Mikkel on April 01, 2014, 04:45:21 am

Comparing SSRC High Precision and Sox 14.4 Very High Quality: http://src.infinitewave.ca/

Lower distortion and noise floor seems to be the only benefits with SOX. However, they are both at or below -130 db so I am not really sure what is gained in practice?
In the other tests at the site SSRC is either as good as or better than Sox.

The SoX VHQ SRC seems to have a lower noise floor, and less potential for aliasing than SSRC High Precision. (I assume Media Center is using high precision)

While something below -130dB should be inaudible on its own, assume that SSRC is 14dB worse than SoX.
That on its own is fine, but what happens when you are also stacking this with other processing which each raises the noise floor by 10dB or more.

Suddenly what should have been inaudible could now have an audible effect on playback.

While the reality is that this should be more of a theoretical change than an audible one, it's possible that it could have an effect. Letting users choose between SSRC or SoX would also let them use SoX's other options such as Minimum Phase. (which may be beneficial when upsampling to DSD)

But it sounds like it would be a lot of work to implement, so it would be low down on the list of priorities for Media Center's development.

theoctavist · « **Reply #16 on:** April 04, 2014, 01:17:59 am »

i have an apodizing filter (matlab) . now if I can find someone who does VST work..

anyone know anyone??

Mikkel · « **Reply #17 on:** April 04, 2014, 01:33:56 am »

Quote from: 6233638 on April 04, 2014, 12:10:17 am

The SoX VHQ SRC seems to have a lower noise floor, and less potential for aliasing than SSRC High Precision. (I assume Media Center is using high precision)

While something below -130dB should be inaudible on its own, assume that SSRC is 14dB worse than SoX.
That on its own is fine, but what happens when you are also stacking this with other processing which each raises the noise floor by 10dB or more.

Suddenly what should have been inaudible could now have an audible effect on playback.

While the reality is that this should be more of a theoretical change than an audible one, it's possible that it could have an effect. Letting users choose between SSRC or SoX would also let them use SoX's other options such as Minimum Phase. (which may be beneficial when upsampling to DSD)

But it sounds like it would be a lot of work to implement, so it would be low down on the list of priorities for Media Center's development.

Certianly I agree that JRiver should have the best of the best. But each of the two resamplers seem to have their benefits and weaknesses. SOX has some phase issues aove 4 khz. I'm not sure how big an issue it is (I'm just a layman). I don't think, though, a noise floor of 130-140db is a problem. Even if you apply some VST-plugins that in worst case lowers the noise floor to 80-90db in most situations that would never be an issue due to the noise floor in the immediate environment. 80-90db is in fact VERY silent.

Anyway, I'm for any progress, so whatever is decided in the future to increase the measurable (although not necessarily audible) quality of audio in JRiver I'm all thumbs up

.

Best regards,
Mikkel

6233638 · « **Reply #18 on:** April 04, 2014, 01:43:00 am »

Quote from: Mikkel on April 04, 2014, 01:33:56 am

SOX has some phase issues aove 4 khz.

Well, the minimum phase and intermediate phase resamplers do - the linear phase resampler is linear.

Quote from: Mikkel on April 04, 2014, 01:33:56 am

I'm not sure how big an issue it is (I'm just a layman). I don't think, though, a noise floor of 130-140db is a problem. Even if you apply some VST-plugins that in worst case lowers the noise floor to 80-90db in most situations that would never be an issue due to the noise floor in the immediate environment. 80-90db is in fact VERY silent.

If it's an issue which can be avoided though, it seems like it would be best to.

39892 · « **Reply #19 on:** April 07, 2014, 04:03:32 am »

Quote from: theoctavist on April 03, 2014, 02:25:50 pm

there are sr plugins i think. r8 brain and r8 brain pro? or are they standalone only.

Standalone only AFAIK. I've yet to find a VST SRC plugin, and was told it was because the VST spec did not allow it.

astromo · « **Reply #20 on:** April 07, 2014, 04:45:35 am »

Quote from: 39892 on April 07, 2014, 04:03:32 am

Standalone only AFAIK. I've yet to find a VST SRC plugin, and was told it was because the VST spec did not allow it.

I hope you don't think that MC is not compatible with VST:
http://yabb.jriver.com/interact/index.php?topic=77399.0
Or is your comment related to that specific plugin?

theoctavist · « **Reply #21 on:** April 08, 2014, 02:23:09 am »

Quote from: 39892 on April 07, 2014, 04:03:32 am

Standalone only AFAIK. I've yet to find a VST SRC plugin, and was told it was because the VST spec did not allow it.

izotope ozone?https://www.izotope.com/en/products/mixing-mastering/ozone/

39892 · « **Reply #22 on:** April 08, 2014, 05:32:04 am »

Quote from: astromo on April 07, 2014, 04:45:35 am

I hope you don't think that MC is not compatible with VST:
http://yabb.jriver.com/interact/index.php?topic=77399.0
Or is your comment related to that specific plugin?

I did see someone mention that JRiver could use VST plugins, although I've not tried any in JRiver. My comment was about the lack of any VST sampling rate converter plugin being available, for use in JRiver or any host.

39892 · « **Reply #23 on:** April 08, 2014, 05:42:41 am »

Quote from: theoctavist on April 08, 2014, 02:23:09 am

izotope ozone?https://www.izotope.com/en/products/mixing-mastering/ozone/

I don't believe Ozone includes a sampling rate converter, at least in the VST Ozone plugin version. I think the Izotope SRC is only in their standalone RX product, or in the standalone Mac program Sample Manager. Also integrated in Sound Forge, but not as a VST plugin afaik. I've looked everywhere for a VST SRC plugin to replace the built-in in a number of hosts, but have found nothing.

theoctavist · « **Reply #24 on:** April 09, 2014, 03:56:55 am »

Quote from: 39892 on April 08, 2014, 05:42:41 am

I don't believe Ozone includes a sampling rate converter, at least in the VST Ozone plugin version. I think the Izotope SRC is only in their standalone RX product, or in the standalone Mac program Sample Manager. Also integrated in Sound Forge, but not as a VST plugin afaik. I've looked everywhere for a VST SRC plugin to replace the built-in in a number of hosts, but have found nothing.

well.....i have the apodizing filter coefficients IF i can figure out how to implement them in a VST plug or something.

mark_h · « **Reply #25 on:** April 09, 2014, 04:00:27 am »

JRiver should implement apodizing. It would be a another sound quality feather in their cap and propel them even further ahead of the competition.

Blaine78 · « **Reply #26 on:** November 15, 2014, 03:44:05 pm »

Quote from: mark_h on April 09, 2014, 04:00:27 am

JRiver should implement apodizing. It would be a another sound quality feather in their cap and propel them even further ahead of the competition.

+1
would greatly appreciate this, if apodizing implemented I'd use the upsampling feature for serious listening. upsampling without apodizing filter sounds smeared IMHO

lisbethfox · « **Reply #27 on:** November 25, 2014, 12:31:52 am »

I'd like to integrate Weiss Saracon's algorithms for resampling, especially what they use for DSD (my preferred is POW3 for most material). While JRiver handles up sampling well, there's no reason I shouldn't be using Saracon's algorithms.

INTERACT FORUM

Author Topic: How does JRiver up sample audio...? (Read 17505 times)