My experience with different music players.
Dec 22, 2020 at 5:41 AM Post #136 of 205
Did a few more experiments. I first tried xxhighend but couldn't get it to work properly in my system.

Since music playback generally improved with full ram buffer based players I wanted to give RAMDisk configs a try. I downloaded AMD RAMDisk and created a 2GB RAM drive. Installed musicbee, PlayPCMWin and albumplayer one by one on the RAMdisk and configured all to use full RAM buffer. Then I tried playing the local music stored in the drive (since I have already ensured memory playback I didn't initially feel like putting the songs to the memory disc). All the players sounded cleaner for the most part (compared to having them installed in hard drive) but also had the 8khz hiss/resonance like PlayPCMWin did in earlier normal disc. It seems fine initially since the remaining frequency spectrum is cleaner and more detailed but gets increasingly annoying as time goes on.

Then I tried another experiment. I moved the songs to RAMdisk and then played them. It surprisingly made a meaningful improvement with a reduction in 8khz hiss and lesser overall haze despite the players already configured to do RAM playback. I have no reason why this happens. But even now the 8khz hiss was quite hard especially when high pitched voices are present in the music. And I preferred musicbee + ram playback installed on local drive instead of ramdisk because of this reason. Apart from the 8khz hiss, in PlayPCMWin the overall higher pitches felt slightly softer, lacking that last edge of bite and bass also felt a tad aliased for lack of a better word (despite the better resolution). This softness was worser in PlayPCMWin installed in non-ram disk so it's an improvement here.

So after everything was done I thought I'll make one final experiment. I uninstalled my apogee dac driver and reinstalled it on the RAMDisk. And now everything made started falling in to place. The 8khz hiss on all the three players reduced drastically and the bass improved a lot. PlayPCMWin is my favourite of the three in this config (settings mostly default, only priority being changed to pro audio thread). There is still an 8khz hiss but it's much milder and occurs only in very few songs. PlayPCMWin also retains a much milder version of that slightly soft tone where the upper registers lack that very last ounce of bite but PlayPCMWin is overall cleaner and less hazier than the other players I've tried. Overall, aside from this occasional 8khz tizz and mild softness it's spatial and resolution properties are excellent now and dare I say, I can hear more detail than even wtfplay!!
All of my massive experiments to come to an understanding for what streaming service was best overall, that’s to say, what was best in most categories via comparison of the service download to SACD and CD material. My conclusion: Apple Music>usb DAC (OPPO-205 up sampling to 24/192)>Sony TA-E9000ES RCA analog input presents music virtually indistinguishable from SACD or CD of same music selection.
 
Last edited:
Dec 22, 2020 at 6:16 AM Post #137 of 205
Not sure why you feel so (reasons aplenty), but if you're Happy that's what matters the most.

For me personally, it has been fairly easy to pick out MP3/aac vs FLAC/wav, and I am not into streaming, even if it were lossless. This entire thread was about sound quality differences between different music playback software all in bit perfect config, arising from different instruction/data transfer sequences and the noise associated with them, and sharing findings with people who experience similar.
 
Dec 22, 2020 at 10:39 AM Post #138 of 205
Not sure why you feel so (reasons aplenty), but if you're Happy that's what matters the most.

For me personally, it has been fairly easy to pick out MP3/aac vs FLAC/wav, and I am not into streaming, even if it were lossless. This entire thread was about sound quality differences between different music playback software all in bit perfect config, arising from different instruction/data transfer sequences and the noise associated with them, and sharing findings with people who experience similar.
Apple Music is AAC digital music media, as well as playback software, formerly known as iTunes; thus, germane to this conversation. So with that in mind:
Here's an experiment I made recently which suggested Apple Music is the BEST player and streamer all manner that best can be discerned: music selection, recording quality, convenience, browser, library, ripping, store, radio, playlists, music player, and much more.

A few weekends ago I purchased Classic Hauser, a high dynamic range recording, on LP, CD, 24/96 FLAC Download, and Apple Music Download. Playing these, nine ways to Sunday as listed below, I had hoped the experimentation would determine what medium/player sounded best.

  • LP>Technics SL-1210GR/Shure V15V (SAS)>Sony TA-E9000ES Pre-Pro Phono Preamplifier input
  • 24/96 FLAC Download>Foobar2000>OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • 24/96 FLAC Download>Foobar2000>Creative Sound Blaster X-FI HD (for usb to S/PDIF conversion at 24/96)>TA-E9000ES optical S/PDIF input
  • Apple Music Download>iTunes>Airport Express (wi-fi to S/PDIF 16/44.1 output)>Sony TA-E9000ES optical S/PDIF input
  • Apple Music Download>iTunes>OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • Apple Music Download>iTunes>Creative Sound Blaster X-FI HD (for usb to S/PDIF conversion at 24/96)>TA-E9000ES optical S/PDIF input
  • Apple Music Download>Network connection to OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • 24/96 FLAC Download>Thumb Drive>OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • 24/96 FLAC Download>Network connection to OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • CD>OPPO-205>Sony TA-E9000ES analog input
  • CD>Rip to iTunes in ALAC>OPPO-205 DAC up sampling to 24/192>Sony TA-E9000ES analog input
  • CD>Rip to iTunes in ALAC>Airport Express (wi-fi to S/PDIF 16/44.1 output) Sony TA-E9000ES analog input
  • CD>Rip to iTunes in ALAC>Creative Sound Blaster X-FI HD (for usb to S/PDIF conversion at 24/96)>Sony TA-E9000ES optical S/PDIF input
Conclusion: all processes yielded a satisfying listening experience, whereby tone, sound stage, and detail seemed so similar as to preclude any revelation to what process/music media delivered the most compelling sound; however, three things did stand out: first, the presence of pops distracted LP pleasure, second, play of downloads via OPPO usb DAC, Thumb Drive, or non-gapless alphabetical order Network connection was inconvenient, and finally, the iTunes play of the Apple Music AAC download via Airport Express to the Sony TA-E9000ES was most convenient, and sounded on par with the CD, ALAC and FLAC media. This makes me question any need for a means to music other than APPLE MUSIC. At any rate, the exercise was a fun activity on a cold and dreary weekend which has kept me indoors. Note, Apple Music to thumb drive was not tested, since I perceived it would have required purchase of the album to permit copy to thumb drive.
 
Last edited:
Dec 22, 2020 at 11:08 AM Post #139 of 205
Apple Music is AAC digital music media, as well as playback software, formerly known as iTunes; thus, germane to this conversation. So with that in mind:
I appreciate your experiments and test methodology but you can be open that there could be others who can discern more differences, and enjoy them. Any ways if you're happy with your setup who am I to judge. This thread on the other hand was to cater to experiments and tweaks for people who benefit from these changes.

Did a few more experiments. I first tried xxhighend but couldn't get it to work properly in my system.

Since music playback generally improved with full ram buffer based players I wanted to give RAMDisk configs a try. I downloaded AMD RAMDisk and created a 2GB RAM drive. Installed musicbee, PlayPCMWin and albumplayer one by one on the RAMdisk and configured all to use full RAM buffer. Then I tried playing the local music stored in the drive (since I have already ensured memory playback I didn't initially feel like putting the songs to the memory disc). All the players sounded cleaner for the most part (compared to having them installed in hard drive) but also had the 8khz hiss/resonance like PlayPCMWin did in earlier normal disc. It seems fine initially since the remaining frequency spectrum is cleaner and more detailed but gets increasingly annoying as time goes on.

Then I tried another experiment. I moved the songs to RAMdisk and then played them. It surprisingly made a meaningful improvement with a reduction in 8khz hiss and lesser overall haze despite the players already configured to do RAM playback. I have no reason why this happens. But even now the 8khz hiss was quite hard especially when high pitched voices are present in the music. And I preferred musicbee + ram playback installed on local drive instead of ramdisk because of this reason. Apart from the 8khz hiss, in PlayPCMWin the overall higher pitches felt slightly softer, lacking that last edge of bite and bass also felt a tad aliased for lack of a better word (despite the better resolution). This softness was worser in PlayPCMWin installed in non-ram disk so it's an improvement here.

So after everything was done I thought I'll make one final experiment. I uninstalled my apogee dac driver and reinstalled it on the RAMDisk. And now everything made started falling in to place. The 8khz hiss on all the three players reduced drastically and the bass improved a lot. PlayPCMWin is my favourite of the three in this config (settings mostly default, only priority being changed to pro audio thread). There is still an 8khz hiss but it's much milder and occurs only in very few songs. PlayPCMWin also retains a much milder version of that slightly soft tone where the upper registers lack that very last ounce of bite but PlayPCMWin is overall cleaner and less hazier than the other players I've tried. Overall, aside from this occasional 8khz tizz and mild softness it's spatial and resolution properties are excellent now and dare I say, I can hear more detail than even wtfplay!!

Tldr for anyone who wants to try this experiment:

Step 1 uninstall your dac driver.

Step 2 install a RAMDisk tool like the one from AMD.

Step 3 create a RAMDisk preferrably of 2 gb or more.

Step 4 install the driver and PlayPCMWin in the ramdisk.

Step 5 (optional, marginal improvement) : move your music into this ramdisk

Step 6 : open PlayPCMWin and play the music.

Step 7 (optional) : try fiddling with priority levels in the settings menu for PlayPCMWin.

Other optional steps: disable windows sounds and usb selective suspend, by following the guide here : https://support.focusrite.com/hc/en-gb/articles/207355205-Optimising-your-PC-for-Audio-on-Windows-10

You can also try your choice of other players like musicbee, albumplayer, etc with similar configurations.
 
Dec 24, 2020 at 11:07 AM Post #141 of 205
Looks like something is severely broken in your system.
Any device with a polling rate of 8 kHz e.g. gaming mouse or so?

Nothing much broken. Stock config of surface book running in battery and usb based dac. Not even a chance of ground loop. All usb asynchronous audio dacs adhering to uac standard today have a polling sequence separated by 125us (which corresponds to 8khz). Generally when the whole system is somewhat noisy (like with foobar2000 or any normal player), this doens't stick out as sore thumb but for whatever reason, with better player configs (like full ram buffer playback: eg PlayPCMWin) though the rest of the sound improves markedly, this 8khz can start to feel more prominent and distinct. The only two software suites that I've come across so far that neither have the digital haze nor this 8khz tizz are Hysolid and wtfplay. Rest either have digital haze, or they are clear but with this distinct 8khz issue.

Also to note that the dacs I've used for this comparison don't show any odd behavior with the general sine sweep measurement tests. More rigorous testing methods may be needed to characterize and measure these changes.

@Sterling2 just a quick question. Which of the players and methods I've mentioned in post#62 in page 5, post #81 page 6, post 102 page 7, post #135 page 9 and post #143 page 10 have you tried? And did you try authentic FLAC copies (and not upscaled MP3)?
 
Last edited:
Dec 30, 2020 at 4:59 AM Post #142 of 205
@Sterling2 just a quick question. Which of the players and methods I've mentioned in post#62 in page 5, post #81 page 6, post 102 page 7, post #135 page 9 and post #143 page 10 have you tried? And did you try authentic FLAC copies (and not upscaled MP3)?
My post on the topic describes an experiment where, for the most part, LP, CD, 24/96 FLAC download and AAC Apple Music download of "Classic Hauser" were compared via all possible means that I have to enjoy recorded music. The exercise left the impression that Apple Music sounded on par to 24/96 FLAC; and, since this on par means to music is inexpensive, convenient, and easy to integrate into iTunes Library, it makes me think my Apple Music subscription is a better value than alternative means to music which all satisfy. That's it. Now, "did you try to authentic FLAC...", my experiment was a grins and giggles event, just something to do on cold weekend while confined to my home. The exercise came about from disappointment with "Classic Hauser" on LP which had numerous pops that distracted me and which lead me to buying the CD, FLAC and Apple Music media to see if one of these other avenues to the magnificent album would better satisfy me than the LP. The exercise revealed that all satisfied me more than the LP, and, sounding similar, I concluded my Apple Music subscription is boss. One more thing I have ZERO need to build any structure in my laptop to deliver music streams which might be perceived as better than delivered from another structure. Why? Because Apple Music sounds on par to stereo SACD, and CD which do not have any characteristic tempered by computer delivery, meaning I do not need to explore computer mods to get a better result from any streaming service since Apple Music delivered via current system structure rivals SACD, and CD. In other words, I do not need to seek a solution to a problem which does not exist.
 
Last edited:
Dec 30, 2020 at 8:54 PM Post #143 of 205
Anyone know how to get PlayPCMWin to play files without opening a completely new PlayPCMWin player every time you click on a file?
I know they say to drag & drop a file to the existing opened player, but can it work with just clicking on a file, I see no option in the settings for this.
 
Dec 30, 2020 at 10:20 PM Post #144 of 205
Anyone know how to get PlayPCMWin to play files without opening a completely new PlayPCMWin player every time you click on a file?
I know they say to drag & drop a file to the existing opened player, but can it work with just clicking on a file, I see no option in the settings for this.

I am unable to understand the question properly. Would be helpful if you can attach images.

What I know of PlayPcmWin so far is that after you open the player, you drag and drop the music into the player. When you press play, all the songs are loaded into RAM and the one you selected is played. This I guess is done to ensure that you get both good fidelity (RAM playback) but also the ability to do gapless playback. Beyond a certain limit, your RAM will be full and you will not be able to add songs to the list.

Edit: I understood your question now. You set the playpcmwin as default player for music and whenever you double click a song it opens a new playpcmwin window instead of overriding the existing one? I'll check with the dev and let you know soon.
 
Last edited:
Dec 31, 2020 at 12:55 AM Post #145 of 205
Edit: I understood your question now. You set the playpcmwin as default player for music and whenever you double click a song it opens a new playpcmwin window instead of overriding the existing one? I'll check with the dev and let you know soon.


Yeah that's what's happening. A small annoyance.

Besides that, I do hear better layering with PlayPCMWin over Winyl, which itself was a big upgrade over Foobar. Thanks for putting me on it.
 
Jan 8, 2021 at 1:16 PM Post #146 of 205
Hi mate,

I get this message when I try to install the PlayPCMwin.

1600512177020.png

Downloaded the installation file from think link https://sourceforge.net/projects/playpcmwin/

Any idea what is the problem?

Thanks/
Looks like the developer is aware of this but unsure of the causation since he couldn't recreate the error in his system. Can you try using the portable version of PlayPCMwin (the one that doesn't need to be installed)?
 
Mar 20, 2021 at 11:30 PM Post #149 of 205
I finally got around to trying XXhighend (version 2.1.0) yesterday, and I feel I should have done this months ago - it sounds absolutely phenomenal. It was a messy set up/configuration process, which i'll get to in the later paragraphs, where I'll share all the downsides.

Let me start with the pros.

1. The sound is fantastic, by far the best I've heard so far. There's two aspects to the great sound. First is the reduction of CPU activity and hence the lower noise.

This player runs on RAM, and has options to reduce background tasks, and also optimized to do minimal data transactions so as to have least noise (I'm simplifying the description of a complex process, which in its entirety is far more content than I could understand in full).

Most players I have tried that didn't do RAM based playback had a layer of digital haze to them, except one case of hysolid, which, without being RAM playback based, was clean but then it also had an artifically dense sound. And when other players were used in customized RAM playback, while the overall noise in most areas of the spectrum decreases, some residuals pick back up at certain frequencies, especially at 8khz, and some kind of artificial softness at times. So the fix isn't a trivial <just push it to RAM and forget>. I described the RAM player experiments in full (for musicbee, albumplayer, playpcmwin) in a previous comment.

Of the music player softwares I have testsed, this is just the second one that I have come across that had neither an immediately obvious digital haze or a coloration (other one is wtfplay), and just the third I have experienced that didn't have a harsh digital haze (wftplay and hysolid are the other two).

This is not to say that xxhighend and wtfplay sound identical, they certainly have differences in sound, but they both don't seem to have any obvious deviations (or probably equally deviated from realism in either ends). XXHE sounds like a direct upgrade for album player, retaining a lot of the general texture/presentation of album player but with less haze and more detail, while WTFPLAY is a different sound one that I have not found an equivalent for.
Of course this would mean little if your dac had some extremely high end isolation, but I believe for a lot of cases, it would make similar differences as discussed above.

2. The above are comparisons with "arc prediction" turned off. Arc Prediction is a built in feature of XXhighend, and it is a custom interpolation algorithm! It is not a sinc filter, and its not even a general linear filter. And turned on, XXHE sounds way better than anything else I have heard reproduced on a digital system, and this alone makes the software worth the price IMO. The best references of an instrument I could relate the real world sound to would be flutes which I am quite familiar with. This is the first time I am hearing the micro textures and tonal gradations, (and the tonguing patterns if any) of the flute being rendered with such realism from a system.

And the same realism extends to anything that has subtle gradients and shifts, the sound of winds, claps, foleys etc sounding far more realistic than I have heard before. The next big improvement is with spatial tracking. Whenever the mix has a panning/rotating object, I feel at much better ease visualizing its entire trajectory without abrupt cuts with this. For anything that is dynamic, and preferrably natural (well mastered synthetic music works great as well), the arc prediction does a wonderful job of filling the points in a way that imo makes best fidelity from a listeners perspective (may not be from the perspective of an APX555).

Yes I have tried Hqplayer, with it's oversampler on my pc. Regarding the digital noise, it sounds as hazy as musicbee or winyl (not surprising as they are based on same library), though I guess you could use a network streamer to mitigate a lot of these noise, so for most people it would be a non issue. However, coming to the oversampling, I even tried some high tap custom oversampled data I got from others using different filters and while it did changes, it never brought anything that would make me use it over the built in OS and Noise shapers in my ESS based dac (Apogee groove, supports upto 192khz 24bit), but arc prediction has done that for me. With any music that has a compressed structure though, I felt I preferred the arc prediction off than ON, as the bass density (or thickness) felt a bit reduced and it sounded bland for lack of a better word. Of course, you don't judge a race car by its ability to run on a bad terrain, but it's quite an easy fix here to just turn off the arc prediction in those scenarios. There is optional choices for custom filter tuning, but I restricted myself to let my dac do its own OS job for that scenario feeding 1:1 data. This is not a knock against hqplayer. I am well aware that it works fantastic on many other systems and for sure would be a better alternative to an oversampling ic on the dac chip for many cases, but there's something quite exciting and realistic about arc prediction that I haven't found elsewhere. Maybe I should try some NOS too someday.

I am generally averse to playback software costing money, more so considering WTFPLAY actually sounded better than almost all the paid ones I have tried, but this is one time where I would say it is worth it. The interpolation algorithm makes a very meaningful difference, and can actually be considered a significant part of the DAC. This definitely makes its way to the top of my wishlist. Now to the problems.

1. It is likely to screw up with your system. Always try on a spare system first. It has an issue with windows 10 on my system where the player removes the background wallpaper and replaces it with black screen.

2. I actually tried to make it run a few months ago and failed, and its the main reason for the delay in trying this tool. The instructions are mostly unclear and the UI is very unintuitive without a proper guide (even with a guide, I would call it barely passable). In the UI menu, To the left top is the settings for Arc prediction ON/OFF, Custom Filtering ON/OFF, Volume control and Volume Normalization ON/OFF. (The red indicator corresponds to ON)

To the middle of the top you have options tagged P, L, S and H. Selecting each toggles the middle window between them. P refers to the playlist, I haven't figured out what L means, and S refers to the settings.

You need to go to settings and configure for your dac in the output for its device buffer size (you can get this from your dac's driver control panel), and DAC settings to choose what sample rate and what word length it supports (mine is 24 bits 192.0 Khz).

At the very bottom of the settings you have options to disable certain additional services if you intend to cut down the noise even more.

Once done with settings go back to playlist to see options at the bottom for play, stop , prev etc. Explore is the command to open file manager to input a music file. Load doesn't open file manager, I don't know it's utility. Clear option clears the current playlist.

3. It is easy to mess up with the settings and get artefacted sound. One area I found it very evident is volume control. XXhighend has a volume control that acts in parallel to your Windows DAC volume control. Simultaneously playing around with both XXHE volume and DAC volume at once (with neither at 100%) is guaranteed to introduce a wierd softness/lack of definition in the sound, as if someone applied a wierd lowpass filter. Voices sound fine but unfocused, while drums and cymbals lack all sorts of shimmer if done this way.
My suggestion would be to set up your dac volume at 100% and use XXHE volume control as it would do volume control + interpolation in one go, which sounds very good without any of the above artefacts.

4. A high bandwidth (high slew rate dac) is recommended as per the Dev, and preferably R2R. Otherwise it is likely to mess up the output of the algorithm. My Apogee groove is actually most likely at the corner of passability (has a little over 100khz bandwidth as per stereophile measuremeasurements) but being a delta Sigma it would in the end doing quite a bit of modulations and I guess I'll not be able to see the full potential until I change my dac.

5. No crossfeed. I would have loved to see a well implemented crossfeed plugin for us headphone users.

6. It costs! I think the price is justified (in fact I consider it a bargain for the effort involved and the improvement it brings) but WTFPLAY and Playpcmwin (or maybe daphile too which I couldnt get to work on my system) would still be what I could comfortably recommend because it is free and sounds great for something that is free.
 
Last edited:

Users who are viewing this thread

Back
Top