My experience with different music players.
Jul 28, 2020 at 11:32 AM Post #92 of 205
Getting this thread back on track...

I've been doing more experimenting since I've gotten new headphone/amp combo. I'm using Raal SR1As with a JOT R now. Every time I make changes to my system I feel like my preferences for setup change. Is this an effect of more accurate equipment showing me what sounds better concretely or is this simply a matter of system synergy? Personally, I think it's a little of both, but I think the Raals definitely show me more of the truth than any headphones I've had previously (heavily modded hd600, verum 1, LCD-X, thx-00 purplehearts, etc.). Anyways, whenever I make a big system change like this, and in this case a huge upgrade, I have to go back and redo many listening tests.

Previously I had decided I liked HQ player/XXhighend best on windows and wtfplay otherwise. I wasn't sure if wtfplay was solidly better or not and wasn't super keen on rebooting to do critical listening. The upsampling in HQplayer still sounds better than without, to me, but I'm finding the impact of that upsampling is less than the improvements brought by wtfplay. I would still love for offline high quality upsampling to be added to wtfplay, but one can only dream. I now much prefer wtfplay to anything else, and think it's worth the slight inconvenience of plugging in a flash drive and rebooting. That switch honestly takes less time than running Fidelizer. I'm still unsure however how a perfectly tuned XXhighend system would sound vs wtfplay. XXhighend I still believe in since it was the first player that really made me think "holy crap THIS is how good pc playback can sound." Buuuut I can't get it to run with it's optimal settings and there are SO MANY parameters to tweak it would take ages to figure out which one of each is best anyways.

So yeah, now I'm just using wtfplay for critical/dedicated listening sessions. I've been able to nail down distinct preferences for settings as well. I wasn't sure what sounded best to me before and tweaking some of the settings just sounded different but not necessarily better, so the settings I was using were essentially stock. My preferences are now very much for large buffers and high process priority. This is kind of the opposite of what is recommended but I messed with it for a while before settling. With everything dialed in the blackness of the background, the solidity of images, and the lack of any annoying HF issues makes it super easy to relax into the music. I both want to turn the music up and I don't want to stop listening, this is how I know things have moved in the right direction.

buffer - 16384
n - 8
prio - 95

These are the settings I'm using now, and I also think it sounds better to use just the wav or pcm player on it's own vs the combined one. As well as not using wtfcui. Now I'll create a playlist or add an album to the playlist, exit the gui, then use the wtfplaylist command to play. This is super quick and easy and again is audibly superior to me. Really interested for anyone else using wtfplay to try out these settings and report back.

PS: Using my own music I've decided I like 24/192 renders vs 24/48 with wtfplay, and while It's worth it for me to do that for my own stuff, I'm not about to upsample every album I want to listen bc that would be a nightmare. 90% of my library is redbook, hence the longing for upsampling functionality.
 
Jul 28, 2020 at 10:31 PM Post #93 of 205
Hey champ, congrats on your purchase. I've seen unanimously positive opinions on the raal, from everyone I trust. And yes they are known to be very revealing.

Regarding music player choices, understand that I made this only to give some experiences on "free" upgrades. Even with my srh1540 I can feel that any software I use, the sound does hit a ceil.

My friends have recommended to move to a dedicated low noise streamer - like allo digione, pi2aes, etc. I would possibly add uptone ether regen to the list. What is happening here with the softwares is that they reduce the noise coming out of the cpu through the USB ports (different polling times etc have a different pattern). There will be something inherent to the device no matter how much you restrict it. You'll just be trading off performance in one corner for improvement in another. Of course the software fidelity can be put in tiers. Something like wtfplay will be superior to winyl, musicbee, audirvana, etc in all ways, but it'll have some issues of itself that you notice only when comparing to something like xxhighend.

The streamers I have said above get the data, and regenerate it in a cleaner, more periodic way. Please avoid any cheap regenerators, especially passive ones like audioquest jitterbug. They do more harm than good. Remember, the data you send is a signal of fast pulses (along with associated noises that join unfortunately) and any normal low pass filtering like those ferrite beads, while reducing noise, will also affect data, putting more stress on the phy which could cause a different type of sound degradation.

I would also recommend you the uptone uspcb cable to connect from the streamer to the dac. Preferably use Ethernet cables to connect from pc to the streamer.

I know there will be skeptics trying to deviate the content without actually trying stuff, but I assume I have convinced you enough with the earlier posts, and your trial and error experiences.

I wouldn't want to spoonfeed on this topic. Rather I'm just going to throw in terms for you to go about and explore at places like sciencedirect - multi layer pcb design and benefits, ground plane noise, phy layer design - power profiles and noise. The uptone uspcb I have is a multi layer pcb with a good ground plane design to minimize noise during data transmission. The ether regen is something that generates the data again, with a good ground plane design to minimize noise at the source in the first place.

Regarding oversampling, your dac will have an oversampling ic inside it that will do the same job. It is needed for proper reconstruction. But more often than not the ic inside is of poor fidelity (low computation power, truncations on the calculation), and this is why custom oversampling from PC helps, since you can do it with a large tap count and hence cleaner reconstruction. Some DACs like chord's stuff might have great reconstruction filters inside them already negating the need of custom oversampling,depends on your dac tbh.

Please do PM me if any further help/advice needed.

Btw I also read that the JOT R had a switch or something that tweaks the sound a little. One of them for a more open sound, and another for a less open but punchier sound. Is it true? Have you experiemented with it if any is there?
 
Last edited:
Jul 29, 2020 at 7:09 AM Post #94 of 205
Just your friendly reminder that the concept of magnitude does exist. And so does the notion of hearing threshold. Manuel doesn't seem to care as he digs deeper and deeper into what starts to look like audio paranoia, but the readers might.
 
Jul 29, 2020 at 9:42 AM Post #95 of 205
Just your friendly reminder that the concept of magnitude does exist. And so does the notion of hearing threshold. Manuel doesn't seem to care as he digs deeper and deeper into what starts to look like audio paranoia, but the readers might.
Yaay, the legend is back in a place he has no business. To save us from our own delusions!! Thank you warrior for fighting against the evils of snake oil and protecting us and our wallets (from free upgrades like tweaking music player softwares).

There is fair postive responses from others who have tried different player softwares on the differences too, who are you really trying to educate here? Anyone who isn't interested wouldn't even peek in here.
 
Last edited:
Aug 1, 2020 at 10:18 AM Post #96 of 205
Hey Manuel, thanks for the detailed reply!

The Jot R does indeed have a baffle compensation switch that makes up for the lack of bass with the OB design. The main weakness with the Raals is the lack of headroom the driver has in the low end before it clips. I listen to a good bit of hip-hop/trap/electronic in which this becomes and issue and I often will not be able to turn the volume quite loud enough for pleasure listening before the driver bottoms out. I have found that by turning the baffle compensation off and then using EQ to bump up the bass in a more gentle manner(less boost + sub 30hz roll off) gives you back a great deal of headroom and I can listen to most any song at satisfyingly loud levels. The Jot R with the baffle compensation turned off and no make up EQ applied can be ok for some songs but it definitely is approaching the territory of treble fetishism. It's too much for most music IMO, although you can basically turn them up as loud as you can handle at that point.

I've been interested in trying out some sort of dedicated streamer/endpoint for a while now, but I haven't yet committed to it. I'm interested in what you have tried out so far and how it compares to pc playback. I'm also curious as to how much improvement there could be over say a custom built wtfplay pc. Like, if I built a fanless pc with a linear powersupply, low latency ram, nice usb card, etc. how would that compare to something like a pi2aes or ultrarendu? And when talking about digital sources, do you think there are other considerations besides eletrical noise and jitter? A HQplayer endpoint could also be interesting to get the benefits of the offline oversampling + low noise. These are definitely things I would like to play with.

I'm using an RME adi-2 pro as my DAC currently and it has an incredibly nice USB implementation. I have a schiit wyrd, which I know is not supposed to be in the same league as a lot of the nice usb reclockers, but with the RME it doesn't seem to improve anything. I'm wondering if the improvements I might get from using a streamer/endpoint would only come from them being lower noise, since the RME seems to do such a good job of reclocking the signal and rejecting jitter on it's own.
 
Aug 1, 2020 at 1:40 PM Post #97 of 205
The main weakness with the Raals is the lack of headroom the driver has in the low end before it clips.
Is it a weakness of the raal or of the jot r? Of course price is a serious consideration, but you can try other more powerful amps and see if things go better. I've seen a lot of dynamics breathe new life with better amps.

I have found that by turning the baffle compensation off and then using EQ to bump up the bass in a more gentle manner(less boost + sub 30hz roll off) gives you back a great deal of headroom and I can listen to most any song at satisfyingly loud levels.
I hope you're using hqplayer convolver or some DAW in high precision for doing the eq. A normal player wouldn't cut it , and would truncate too much.

Alternatively you could download reaper and try some compression plugins. People look at compression as something bad, but too open sound can be unengaging for edm and compression can breathe life into it. Some headphones like thx00 are already compressed sounding and hence work fine with them, with the flaw that they can never sound open when called for (binaural, acoustic etc). On the other hand, something like a hd800, while sounding too open stock can take in compression and add density to the sound as you like. Apart from compression, you can also experiment reverb. These non linear operations add lot more to sound than just an eq can ever do.

I'm using an RME adi-2 pro as my DAC currently and it has an incredibly nice USB implementation. I have a schiit wyrd, which I know is not supposed to be in the same league as a lot of the nice usb reclockers, but with the RME it doesn't seem to improve anything. I'm wondering if the improvements I might get from using a streamer/endpoint would only come from them being lower noise, since the RME seems to do such a good job of reclocking the signal and rejecting jitter on it's own.

The fact that you hear changes in music player software says that there is somehow noise seeping into the system with changes in source chain. I would recommend AGAINST cheap regenerators, since they could do more harm than good. Just putting a rc or l filter of sorts (like ferrite beads) while controlling RF noise would also blunt the actual digital data which is also in the same frequency. You'll at best be moving from one aberration to another.

I personally don't have much money for a dedicated streamer but the uptone uspcb cable helped quite a bit in my system. The design is genuine, it's just a pcb with good ground plane design to sink out the switching noises when the high frequency data pulses pass through. And the designer is a cool guy which is why I'm fairly positive that ether regen could be good since it'll now also reduce noise from the input side as well.

Definitely try an Ethernet based streamer to see changes. Don't jump full force on first try. Maybe just give raspberry pi4 and a decent power supply configured to do volumio/any other streaming os first to see if it improves. Then you can start moving up the order.

Regarding comparison to pc, it'll surely be better than a normal pc. But a pc configured just for audio? I'm not sure, but they are too expensive. You'll be upgrading all the power supply etc inside it for that.

The holo may has a very good usb isolation and even then it apparently benefitted from a streaming setup.
https://www.head-fi.org/threads/wha...igital-audio-from-your-pc-to-your-dac.932199/
 
Aug 1, 2020 at 2:19 PM Post #98 of 205
Is it a weakness of the raal or of the jot r? Of course price is a serious consideration, but you can try other more powerful amps and see if things go better. I've seen a lot of dynamics breathe new life with better amps.

It's a limitation of the ribbon transducer. Has so many good properties it has to give a little bit somewhere. The Jot R has a crap ton of headroom, at least 20+db even for classical/jazz.

I hope you're using hqplayer convolver or some DAW in high precision for doing the eq. A normal player wouldn't cut it , and would truncate too much.

Using the PEQ built into the adi-2 pro.

I'm definitely going to try out a pi streamer soon and see how it goes. I've been torn for which path to go down when I build a new pc for music production. Building a daw pc and building a transport pc have rather competing design goals. If I could just build a pc for production and the listen via streamer that would certainly be the most elegant solution.

Still interested if anyone else sees any benefits from the wtfplay settings I posted.
 
Aug 1, 2020 at 2:27 PM Post #99 of 205
I'm not sure of RME stuff, but precision eq needs quite a bit of processing brute, it is unlikely the one inside rme would do as good as a DAW based eq. The proceeding brute required is about as much as it is required for oversampling with high precision (chord has an entire FPGA unit to just to oversampling for reconstruction, and people with hqplayer offload tasks to GPU for similar tasks). They are essentially similar tasks after all. Definitely give any daw quality plugin or design a high precision fir eq for hqplayer (I'll try to make one if I can) for best fidelity. I know it sounds pedantic, but atleast reaper can be tried with ease.

Regarding buffer settings, it's extremely system dependent. Even changing my usb cable forced me to tweak certain settings again.
 
Aug 1, 2020 at 3:59 PM Post #100 of 205
The RME has an FPGA dedicated to fx. The built in EQ is very nice. Of course, if I could run something like DMG Equilibrium that would be even nicer...
 
Aug 1, 2020 at 10:56 PM Post #101 of 205
That's pretty nice. 😊.

Do let me know if you get to try a sbc streamer. I have an allo sparky at my house but unfortunately I don't have a decent power supply to go with it. Hence haven't tried going the streamer side yet. Was about to try and build a portable set with batteries, but then life got a bit busy. I'll let you know if I get to set it up.

Also try some compression/reverb plugins if you're interested. Extremely clean devices sometimes subjectively benefit from experimenting there. Besides, those are an integral part of mastering.
 
Sep 18, 2020 at 11:18 PM Post #102 of 205
you should try WASAPI in uLilith :)
I agree Foobar is not very good for fidelity. Even Winamp does a better job. Here is my current top music players downloadable for free and the first one does a spectacular job!

1. JetAudio
2. Winamp/Zoom player
3. MHC player
And then you have many including Foobar indeed. Not going to bother at this point.

Go at oldversion.com and download the popular version of JetAudio. Give me some feedback what you think. By the way I have asio/wasapi uninstalled.
Try "Russian Album Player" or "PlayPCMWin"
I like winyl and ulilith.

I have finally got around to trying PlayPCMWin, Smallplayer, Ulilith and JetAudio. There is a change to my system during this analysis. Earlier I used my apogee groove + HD800, or apogee groove + Burson fun (sparkos ss3601 opamps) + SRH1540. PC remains same - surface book. This test was done using topping nx4dsd, and urbanite xl, both of which are brighter, less resolving than the formers, but I've tried my best to decouple this from the music player effects. I also did the USB selective suspend removal, and removed system sounds as per this guide from focusrite : https://support.focusrite.com/hc/en-gb/articles/207355205-Optimising-your-PC-for-Audio-on-Windows-10 . Buffer size on my dac was set at 32768 samples, the max permissible value, on all scenarios except Ulilith (reason below).

1. JetAudio - Unfortunately it doesnt seem to support asio/wasapi. And sq was only Okay. I'm not even sure if it supported usb asynchronous audio. Enhancements were turned off. I'd be happy to give it another try if anyone can guide me to set this up in bit perfect Asynchronous audio.

2. SmallPlayer - Sq in terms of texture was alright not much problem with unnatural shimmer or bad boomy or bootleggish bass, but noticeably mistimed. Pans, etc weren't so great. Not sure if it supports usb asynchronous, I doubt it considering how old it is and the sound, I seriously doubt it. I guess it would be a very nice pairing with internal cards on pcie or otherwise. I did try it on my Laptop speakers directly and it performed quite well there, but a bit quiet and extremely lacking in bass. (I guess showing the speakers with their actual properties instead of faking density using compression or other artefacts. It had noticeably more energy and speed than I'm used to with my laptop speakers). I'm gonna keep it for my laptop speaker playback.

3. PlayPCMWin - Very nice. I like it a lot. One of the few players that compete well against musicbee in my setup. Had to fiddle a little with settings. Settled on Exclusive mode with Pro audio priority. 16bit downscaling I left to truncation instead of other options, but it's not a deal since my dac accepts more than 16bit data streams. In terms of sound reproduction, while it is noticeably different from musicbee, neither of them feel coloured or wrong. It's the presentation, and texture around instruments where the differences pop in. PlayPCMWin sounds a lot more efforless compared to Musicbee. Musicbee has a bit of an edge to the sound and as if the instrument had an accompanying low level noise band, in comparision to PlayPCMWin where the instruments sounded fairly natural without any additional band/texture around them. However I feel that there is a problem with that too, PlayPCMWin seems to shave off the edges to the sound of instruments a little. I find transient hits to be slightly more impactful and real despite the slight noise around them on musicbee, while PlayPCMWin seems to mute it just a tiny bit away from what could be perceived real. It is tiny, but it is noticeably and makes me wanting a little more attack. It also makes bass guitars and low level echoes harder to follow than Musicbee even though it is cleaner on that aspect. I can sum it up this way. Musicbee shows good detail and hints of what could be hidden extreme detail, but is a bit distracted sounding. PlayPCMWin shows good detail in better precision but doesn't try to extract the hidden extreme details like musicbee does. PlayPCMWin is the closest I have gotten from windows to my current reference WTFPLAY linux distro, in terms of texture, effortlessness and most importantly lack of color. Hysolid is more detailed than PlayPCMWin and also is just as effortless while extracting a lot of shades of shimmer, but is more denser sounding than what could be perceived real. It's a bit of toss between hysolid and PlayPCMWin, but hysolid is buggy, PlayPCMWin is not. WTFPlay is still my reference/champ, it has the texture, and effortlessness of PlayPCMWin but also far more detail, better realistic impact, and believable sound than all of PlayPCMWin, Hysolid and Musicbee.

4. Ulilith - Seems to have issues with large buffer sizes, was stuttering heavily on 32768 samples on the buffer so I dropped down to 2048 samples. Works great in Wasapi. It is very different from PlayPCMWin and Musicbee. Both the star and pitfall of the show is the treble. Its sizzly, shimmery, but also aggressive. It is not distorted or fuzzy. It is very real, but very forward and in your face. Like you are standing next to a cymbal crash, with no muting off of any part of the sound. Oucch. My current setup being a bit treble fuzzy doesn't help it much either. I expect it to be a lot more enjoyable when I get back to my Srh1540 or HD800. It veers on the opposite end of PlayPCMWin if you keep Musicbee in the middle. Both Ulilith and PlayPCMWin sound cleaner than Ulilith, but both show deviations in either directions. While PlayPCMWin is mellow and laid back, Ulilith is forward and aggressive but without any of the general distorted feel which i get when eq'ing up the treble. I love the energy and shimmer of Ulilith, but I would have to be very careful of system pairing. One more point to note, I first tried buffer to Auto from 32768 which was unbearably aggressive during cymbal crashes. Setting it to 2048 made it more bearable, I think fixing it to a more optimal value would help me get it to good enjoyable levels while still maintaining that aggression.

I would be experimenting further on both PlayPCMWin and Ulilith with my other gear, as of now, I like PlayPCMWin a lot and Ulilith as a flavor player. Musicbee still remains my go-to choice due to it being the middle of the line, despite its minor issues, but all 3 are compromises in one way or other, when you compare to something like WTFplay.
 
Sep 19, 2020 at 2:12 AM Post #103 of 205
HQplayer would push human sound in front of you and then eliminate all the other background sounds. That's the most failure problem, Sinc filter not always right. you can not just digitalize all the recorded audio samples, and then treat them to Fourier. At least, Foobar 2000 gives you the most authentic enjoyment
 
Sep 19, 2020 at 2:32 AM Post #104 of 205
HQplayer would push human sound in front of you and then eliminate all the other background sounds. That's the most failure problem, Sinc filter not always right. you can not just digitalize all the recorded audio samples, and then treat them to Fourier. At least, Foobar 2000 gives you the most authentic enjoyment
I'm unsure how you came to that conclusion. No filter is right because we have chopped off the information anyway, and Nyquist condition is a limit tending to infinity condition. Reality is we window it and different windows are technically different. But chances are, the dac you have also has a sinc oversampler, and often likely to be less precise than the ones generated through hqplayer. Also hqplayer does not have just one sinc filter, there is flexibility to choose a lot of filters depending upon your preferences.

Sadly, foobar2000 doesn't give me any sort of enjoyment, let alone an "authentic" one. My piece of cake is wtfplay. You can have yours.
 
Last edited:
Sep 19, 2020 at 3:55 AM Post #105 of 205
Edit:

I went and edited my complete post. Really some of this stuff has always been rehashed over and over. Maybe there is some form of entertainment pushing a set of audiophile values. We must all be this way? Hah
A7FF939B-F7A9-4FCA-AC21-860DBB1813BC.jpeg
92FC67E3-2B93-40DE-9BDA-81788B30832F.png
A28F48F1-60B3-40C6-832C-AADF6BF78F00.png


https://colibri-lossless.com/themes/

Probably there could be many choices and ideas as far as what’s best. I don’t use a PC for music playback but used Foobar2000 for the last 10 years off and on. Probably the greatest thing about Foobar2000 is the ability to quickly transcode files over. Say you want to listen to a 24bit FLAC on your phone but want to seamlessly integrate the file with your iTunes? Foobar2000 after all the available needed codecs downloaded can do just that. Quick convert to a file on the desktop and you actually have a hi-res file in ALAC you can play both on your phone or on your PC.

I’ve recently switched to the $4.99 Colibri on the Mac as it sounds great. They make Foobar2000 for the Mac but it’s a simplistic idea and nothing like the full PC affair.
 
Last edited:

Users who are viewing this thread

Back
Top