The Road to JPlay

Software Type: Music Server Software for PCs
Price: €99.00

Apple recently released an update for Mountain Lion 10.83 that included Boot Camp support for Windows 8. I have a copy of Windows 8 Pro 64 bit that I thought would work well with my Early 2011 MacBook Pro. This MacBook Pro is a 2.3 GHz Quad Core i7 with 16 GB RAM and a Samsung 840 Pro SSD 256GB. My music library is on a Promise Pegasus 8TB drive and an exact copy on a G-RAID 8TB Thunderbolt drive.

Installation of Windows 8 with Boot Camp is easy and was uneventful. Complete instructions for this installation in Mountain Lion OSX 10.83 can be found here:
I was unable to use the Pegasus Thunderbolt drive as there are no Windows drivers. The G-RAID was supported in Windows 8 so I was in business. My drives are formatted with HFS+ and the files are AIFF. so I needed a program to allow Windows 8 to read and write to the drive. I used MacDrive Pro:
This program worked in the background allowing Windows 8 to access the G-RAID drive.

I then purchased a copy of JRiver Media Center 18 as my music player. JRiver imported the library with no issues and was easy to set up. The sound of the system with JRiver was very good playing everything from 44.1/16 to 192/24 and native DSD 64 files.

Additional Windows settings included turning off Windows Update, turning off Pagefile, System Restore, and turning off allowing the computer to sleep.

I used the excellent remote program JRemote for the iPad that did a first class job controlling JRiver.

The DAC used for evaluation was the MSB Technology The Analog DAC.

I have read numerous reports from audiophiles using an add-on program to JRiver called JPlay. JPlay is the creation of two dedicated computer audiophiles; Josef Piri and Marcin Ostapowicz. What the JPlay plug-in does is substitute its own sound engine for other software music players such as JRiver or Foobar. JPlay claims that their software will improve the sound of programs like JRiver or Foobar, but retain the library functions and other features of these music programs.

What is JPlay?
What does JPlay actually do to improve the sound of already excellent players like JRiver? JPlay improves the way your computer deals with memory playback, optimizes CPU functionality, and prioritizes essential computer functions while turning off non-essential threads.

The JPlay player is small enough to load into CPU cache and can be used alone or with other software players. JPlay provides optimization of the playback buffer resulting in very low latency to improve playback performance. The large page memory used in JPlay results in low CPU latency as a result of superior memory management. JPlay is able to access non-fragmented memory sooner than other programs as it works as a Windows Service. Music data playback is uninterrupted by giving it the highest maximum priority. JPlay supports both 32 bit and 64 bit CPU architecture.

JPlay supports 16-bit Red Book and high-rez 24-bit files in WAV, AIFF, ALAC or FLAC formats as well as native DSD 64 and 128 files. As I mentioned before, JPlay works with JRiver Media Center, foobar2000, MediaMonkey, music streaming services: Spotify (via fidelify), Qobuz and many more—the only requirement is that the host player supports ASIO output. JPlay also provides JPlaymini; a stand-alone player that works by copying track links from programs like JRiver, or other players and pasting them into Mini. JPLAY also supports a StreamerMode (setup with two PCs connected via LAN) with a dedicated ‘Audio PC’ fully optimized for music transport.

JPlay works with Windows 8/7/Vista, but not XP. 64 bit Windows 8 provides the best results with a Quad Core CPU and 4GB RAM (2GB is minimum). As you can see, JPlay’s approach likes a fast CPU with large amounts of memory.

I have just touched on the many features and functions of this software. The JPlay web site will provide even more information as to what this software is actually doing.

Now some of you will ask why bother with a program like JPlay when programs like JRiver are bit perfect unless DSP options are employed? The answer rests in the fact that music playback is more than just reproducing a data file. It has the element of time involved in the process. How the computer deals with the time factor can have a profound influence on the quality of sound that plays though our systems.

I will share my experiences and results of using JPlay with JRiver in Windows 8, and how this combination sounds in comparison to OSX programs like Pure Music and Audirvana Plus.

JPlay Settings
If one really wants to get the best from JPlay, you should set aside a period of time to try the different settings and music engines. I worked with the program for two weeks before I found the optimum settings that sounded best to me. JPlay offers 4 different playback engines. ULTRAstream requires Windows 8 for an audio single PC setup like mine and demands the use of Kernel Streaming.

The Playing Via box offers Kernel Streaming, WASAPI, and ASIO. Kernel Streaming offers much lower latency and is the lowest audio engine level in Windows, allowing more efficient, real-time streaming. For the ultimate playback experience, JPlay recommends Kernel Streaming. ASIO is applicable only to DACs that have an ASIO driver. Note that on 64bit Windows a 64bit ASIO driver is required.

There are 3 buffer settings. The first is for 44.1 files, the others for 96kHz and 192kHz files. The best selection would be DirectLink. The lowest settings will result in the best quality playback. I found that I was able to select DirectLink in the Buffer and Buffer 96kHz, but had to use 2 samples to avoid issues with the 176.4kHz and 192kHz files.

Throttle: [OFF,ON]. When Throttle mode is activated, JPLAY will reduce to absolute minimum both CPU and I/O priorities of all running processes and individual threads in the system it can get to, without risking the stability of Windows.

Hibernation Mode: This mode eliminates dozens of OS jitter-inducing processes & hundreds of threads, providing the best possible sound quality. This is an extreme setting and was not used in my single PC setup and is probably best utilized in a 2 computer streaming setup.

There are also settings one can tinker with in the Registry. The only one I used was the UltraSize setting that controls the size of the buffer for the ULTRAstream engine and ultimately settled on 25.

Setting up JRriver for JPlay couldn’t be easier. Select ASIO in Audio Output Mode. Select JPlay Driver in Output Settings.

The Sound
I first compared the sound of JRiver with WASAPI Event Style and the MSB USB driver for The Analog DAC with JRiver using the MSB ASIO driver. It was no contest. Running JRiver in ASIO sounded much better to me than WASAPI; a bigger soundstage with superior detail and focus. I was very impressed with JRiver ASIO.

As good as JRiver sounded, JPlay in ULTRAstream was superior. The soundstage was larger in both depth and width. Bass was tighter with more impact and far more dynamic. What really got my attention was the superior clarity, focus and definition with JPlay.

Aaron Neville’s My True Story 192/24 showed these differences easily. The bass had pop and punch with JPlay; without JPlay it was good, but the dynamic pace and rhythm was not as good when using JRiver with ASIO. The soundstage opened up with JPlay with the background vocalists clearer and more easily heard. The sonic stage seemed compressed without JPlay. Aaron Neville's voice just sounded clearer and more focused to me with JPlay. There was no brightness or warmth added to the music with JPlay.

Listening to Patricia Barber Smash 192/24 resulted in the same findings; JPlay was superior to JRiver with ASIO. Soundstage, detail, focus, and dynamic pace and rhythm sounded better with JPlay with Kernel streaming and the ULTRAstream engine.

I would like to comment on the new version of JPlay 5.1 vs. the former version 5b. 5.1 was superior sounding to me especially with the new ULTRAstream engine. The Xtream engine worked much better in 5.1 without the delay found in 5b. I found 5b to be somewhat bright sounding, but this brightness was not heard with 5.1.

I was now curious to compare Audirvana Plus Direct Mode / Integer in Mountain Lion 10.83 with JRiver ASIO and JRiver with JPlay. OSX optimizations were applied such as turning off Spotlight for the external hard drive, turning off OSX auto updates, turning off the infrared remote, and turning off Bluetooth.

Audirvana Plus in Direct Mode / Integer was generally comparable to JRiver with ASIO. Audirvana Plus was a little darker sounding and not quite as detailed. Soundstage was similar for both programs. I preferred the slightly more open sound of JRiver ASIO and felt it was the more revealing program. I have no doubts that some of you will like the warmer presentation of Audirvana Plus, but I never felt that JRiver ASIO was bright sounding.

Audirvana Plus and Pure Music did not measure up to what I heard with JPlay. JPlay just stole the show from these two programs.

I encountered a problem playing native DSD files with JRiver/JPlay. It just wouldn't work no matter what combination of settings and engines I tried. JRiver with WASAPI Event Style and the MSB USB driver did play the native DSD 64 files that I had. I was able to play native DSD using the JPlaymini program with no issues. And yes, it sounded excellent!

JPlay 5.1 with JRiver in Windows 8 Pro 64 bit is now my reference player. Marcin and Josef have done a great job with JPlay. I suggest that you give the trial version a listen to see if it improves the sound of your system as much as it did for me.

JPlay Update 4.21.12
I have received numerous questions concerning proper setup of JPlay. My advice is to concentrate on getting the lowest settings for the 3 buffers. DirectLink is optimum and will provide the best sound quality. You can then address the registry settings such as the UltraSize for the UltraStream engine.

Hibernate Mode
I didn’t say much about Hibernate mode as it won’t work for everyone. Hibernate Mode stops many computer processes to reduce computer “noise”. If you want to try it, insert a USB flash drive and start JRiver or the JPlay mini player. When you are done listening, remove the USB flash drive and the computer will be accessible again.

I did not experience a big improvement in sound quality with Hibernate Mode. A slight increase in clarity was observed. Not a big enough difference to give up the JRemote function that will be lost with Hibernate Mode.

JPlay/JRiver Playing DSD files
I was able to solve the issue of JPlay/JRiver not working together to play DSD files with the MSB Technology The Analog DAC. I uninstalled the MSB Technology drivers and reinstalled them. JPlay/JRiver now functioned with both DSD64 and 128 files.

I’m looking forward to hearing about your experiences with this excellent program.

Associated Equipment

valenroy's picture

Hi Steven,

Thanks for the writeup and it is an interesting read on the JPlay as usual but for 99 EUROs on this add-on player to achieve greater sonic fidelity based on the idea of switching off non-critical background processes, why not try Fidelizer (with 3 preset levels of background-processes optimization and a custom mode). I had tried both and I personally prefer the Fidelizer with my subjective taste in audio fidelity. Anyway, it is a free download and you might be surprised at what a free software can provide!

p/s: You can also try both Fidelizer and JPlay simultaneously too!


robertjwarren's picture

Mainly for other programs, but I am curious why you would use windows for music on a mac as I have always heard that mac's perform music stuff better than windows. 

Steven Plaskin's picture

Thanks for the suggestion Ivan. JPlay provides 4 sound engines that do best with more computer processing power. Fidelizer does not offer this.

I agree that both programs should be tried by potential users. Fidelizer's price is hard to ignore.


JRand's picture

There have been tests done showing that JPlay outputs the *exact* same waveforms as stock JRiver. A detailed analysis can be found over on the hydrogenaudio forums. Basically, if you have JRiver (Windows/OSX) you already have bitperfect playback, and JPlay will not add anything to the sound. With everything digital, it is trivial to capture what is output down to the tiniest detail, and then compare them side by side.

They have also tested Windows JRiver vs OSX JRiver. Same exact waveform outputs. (As they should be)

Steven Plaskin's picture

Yes, both programs are bit perfect, but they sound very different due to differences in the time domain in playback. All one has to do is to listen.

JRand's picture

Just listening can be entirely subjective. Moving the angle of your ears in relation to your speakers can make everything sound different.

However, analysing direct waveforms from the outputs directly before they come out of  the speakers cannot be subjective. It is what it is. If A and B output the exact same wave form, the only differences you are going to hear when it comes out of the speakers is from the placement of your ears and items in the room.

Steven Plaskin's picture

I obvioulsy don't embrace the Hydrogen Audio philosophy of "bits are bits". When I can't trust what I'm hearing, it's time to move on to something else.

What I never, ever hear from you guys, is that that you actually tried to listen to what we are discussing.

A real audiophile enjoys music; not waveforms.

hotsoup's picture

gefski's picture

Yep, bits is bits. In the digital domain, power supplies, cables, file delivery, and optimizing the pc for handling music files shouldn't matter. But all of these matter, and cumulatively the sonic benefits of doing everything right adds up to a more transparent window on (and more enjoyment from) the music.

Actually the bits comment states that misunderstanding about high fidelity perfectly "...Jplay will not add anything to the sound...". Thank you! Exactly what we don't want in high resolution audio is to "add to" the sound.

I've been using Jplay Mini (don't care about iTunes or JRiver graphics) for over a year and it's as advertised. A bit too much "fiddle factor" when setting up, but once you've got the settings done for your system, it's great.


Pale Rider's picture

Yes, both programs are bit perfect, but they sound very different due to differences in the time domain in playback. All one has to do is to listen.

I actually believe that everything matters when it comes to such stuff. No matter what one changes, it can have an impact, even though there are times when many of us cannot hear it. Still, I looked high and low on the JPlay website, and while I saw repetitive explanations of getting everything out of the way but sound quality and performance, i saw nothing that would explain how JPlay does anything to the time domain of the data that isn't already being done superlatively by the excellent Analog DAC. Such claimed differences seem unlikely given JPlay's approach to memory buffer management. Indeed, FAQs #5 & 6 suggest exactly that. The argument seems to be: "your DAC can be better, because our software makes its job easier." Perhaps, on a lesser DAC, but on any DAC that can stay the heck away from USB, or can actually completely separate the data from the time domain, count me skeptical.

Not, i am not saying impossible; just skeptical. And while it would be interesting to try, the attarctiveness of setting up a Boot Camp machine just for this purpose dwindles. Perhaps when my Baetis gets here.

Anyway, interesting review. Thanks Steven.

Steven Plaskin's picture

Pale Rider,

You have raised some very good points. The perfect DAC should not be influenced by different USB cables or computer software. But they are. Even the Analog DAC.

Thanks for your comments.

bsm's picture

Pale Rider, you probably need some of these:

Wavelength's picture


Look if you have been in audio long enough then you know everything changes the sound. This is the same with applications, cables, computer types etc...

Also in talking to Steve many times during this review, I will say there is a ton more setup envolved to reach this state. Windows has always been that way and even harder under Windows 8 which they are trying to hide as much as possible. Also there is some major problems with Windows 8 on 2 of my 4 machines which yield them unusuable.

But think about it this way. iTunes is bit perfect if setup correctly and it sounds drastically different than other applications.

More importantly is this... this is software, it is relatively inexpensive and evolving so the investment is small compared to the results. Also try it and see before putting out the cash. You may not like it as much as Steve.



bsm's picture

"iTunes is bit perfect if setup correctly and it sounds drastically different than other applications."


hirezaudio's picture

Steve, thanks for your review and impressions of it compared to your Mac OSX system.

I'm a long-time Mac user and know very litte about Windows, especially using it for audio. But I'm not married to any one OS. All that matters to me is how it sounds, so I'm willing to try JRiver/JPlay and compare it to my current standard (Audirvana Plus).

Before I move forward though, how much HD space does it take to run Windows 8 in Bootcamp?

I have the same Macbook Pro as you, 16GB of RAM and a 240GB OCZ Vertex 3 SSD.

Thanks in advance!

Steven Plaskin's picture

Hi Gary,

I let Boot Camp divide my SSD in half. You don't need that much space, but if your MBP is dedicated, that would be nice. 60 GB would work easily.

Is your music on an external HD? if it is HFS+ / Aiff you will need MacDrive Pro as well. This can be used on a trial basis. Be prepared for a few setting changes after install. Google most of your questions, and you will find the answers. If you get stuck, ask here.


hirezaudio's picture

Thanks Steve for the info about Boot Camp.


hirezaudio's picture

Hi Steve,

I was wondering if you've had a chance to do any more comparisons with the 1.5 Beta version of Audirvana Plus. In standalone mode (deselect iTunes integrated mode), this version sounds absolutely amazing. To my ears, this version is a significant upgrade. I will still most likely install Windows 8 to satisfy my own curiosity, but man... what I'm hearing now is in a whole new league compared to what I was hearing before.


steveoat's picture

Remember how terrible the inital CD players were.  We  were told that there was no differences between players because bits are bits.

We've come a long way.

I have jplay and audirvana and jplay sounds better to me.  JRiver is OK, but not in the league of either.

Listen to your systems, not what people tell you about how it should sound.

coreaudiotechnology's picture

A digital signal is not "bits". In reality, bits don't exist. The digital signal, whether from a computer, CD Player, or network player is a square wave, which is an analog representation of a square wave.

This square wave has both amplitude and timing characteristics. Any timing errors directly become amplitude errors to the circuits doing the digital processing. Noise, jitter, and other forms of distortion create amplitude distortion on the square wave, which translates to harmonic content being introduced into the audio signal that shouldn't exist.

A computer isn't just some brain that is processing ones and zeros. It's made up of billions of transistors all switching on and off based on program logic. The more things that touch the square wave the more noise gets picked up before the output.

The goal of these playback softwares is to optimize the computer's ability to get the audio square wave from software to output as quickly as possible. To do this they prioritize and eliminate background services and dedicate additional threads to audio applications. In doing this they both shrink the signal path to the output and improve how quickly the signal gets there.

The different pieces of software all utilize different algorithms that process the audio data. Any process done to the audio signal creates a new version of the square wave from voltage in the power supply. So you end up getting copies of copies of copies as the signal travels from input to output. Better power, less copies, etc translates to less amplitude distortion on the square wave and better sound.

Some softwares will eliminate more than others. Some will utilize different algorithms or DSP processes. These minute changes alter the resulting harmonic content that is introduced into the square wave and thus the resolution depth of the sound you hear.

There are both scripts available and optimizations that you can do your self to improve upon what jplay does and improve your audio without using jplay. JPLAY doesn't fix computer audio like it claims, the hardware still plays a significant role in the resulting sound more than anything. But optimizing the operating system does influence the sound as they claim.

How much of a difference each piece of software makes will depend on your hardware. The better your power and the faster your computer can process data the less of a difference you will hear between pieces of software. Clean power means clean copies and faster processing means the data exits the system as quickly as possible. Latency Monitor is a great app for seeing how long it takes processes to reach system output.

You can read more information about how the CPU processes data here: along with some other articles regarding computer audio.


highstream's picture

In the list(s) of JPlay's sonic characteristics the review doesn't mention anything about tonal accuracy, which I would have thought would be the starting point for any audio component or software review. As far as I can tell, the Ultrastream setting creates the most processed sound, which depending on the components in the chain may or may not be so great. I've been auditioning ICs (RCA, unbalanced) - MAC/RAC Silver Quad +, Darwin and Grover Huffman - and have found the Ultrastream engine (latest JPlay, with foobar2000) works well for one of them, exagerates the slight brightness of another, while the third sounds overly processed and bright.  The other engine choices bring the tonality closer to earth, even if with changes in depth, texture staging and volume, the latter considerable.  It'd be interesting to see the frequency curve of the different engines.

Btw, one recording I find to be a good test of tonal accuracy with the human voice is Laura Nyro's 1971 Live at the Fillmore album/CD ("Spread Your Wings and Fly").  She had a bad cold that night. One can tell a lot about components, cables and JPlay settings by how they render (or disguise) it.

Using that Laura Nyro recording and a couple of others, including a big band CD, I just compared foobar only vs JPlay via all engines with the Grover Huffman IC and the no-JPlay mode was better in a lot of ways: more natural tonality of instruments and voices, notes sounding like they were played vs. a layer of "processing" masking (or softening) their presence.  The width, height and depth is still there, but there is also a little less distinction between individual notes (e.g., listening to Beethoven piano sonatas).