BDP-2 and 3 and Roon

0 Members and 1 Guest are viewing this topic. Read 3625 times.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #20 on: 3 Apr 2018, 06:57 pm »
@zoom25, you should receive some sort of subjectivist merit badge for all of that testing and reporting...

However, I have to question the part in BOLD, above.

I am too lazy to search through my posts and find the actual numbers, but I've run tests while playing WAV, AIFF, FLAC, ALAC files, looking specifically at CPU utilization on the BDP-1, which I know you have.

Even using 192/24 files, playing from a USB flash drive, the CPU utilization difference between FLAC and WAV was minor and within the same order of magnitude. The only codec that really was an outlier was Apple Lossless (ALAC). ALAC suffers from the flexibility of its meta data storage - you can arbitrarily place meta data just about anywhere in an ALAC file, which makes decoding relatively expensive. Or so I've been told... I've never written an ALAC decoder nor have I looked at anyone else's code.

In my tests, ALAC was substantially higher in CPU utilization. But WAV, AIFF and FLAC were pretty close - yes, FLAC was higher, but the machine was still hovering around 90% idle during playback. I struggle to believe that even an order of magnitude greater CPU utilization will create an audible difference; especially given that the BDP's only job is digital transport via the sound card (again, with a BDP-1).

I know that folks like to ponder the differences between WAV and lossless compressed codecs - even going so far as to recommend ripping directly to WAV vs. FLAC, as even that supposedly creates audible differences. But the computer scientist in me knows that is wishful thinking (expectation bias). Whether or not you encounter an error during the ripping process, saving the data into a WAV container or lossless-compressing and saving into a FLAC container will do NOTHING to change the resulting PCM. I know this because I rip with XLD using the AccurateRip database. Whether I rip to WAV or FLAC, the resulting CRC check tells me that the PCM data is an EXACT match with the original content.

Anyway... enjoy the hobby.

Hello Krutsch,

Yes, I remember your numbers and posts on the CPU percentage on the BDP-1 with different formats. I've done the same on BDP-1 and on Mac with Audirvana Plus and other stuff. We've both monitored numbers. Like you I also use XLD for all my transcoding. The resulting data is always the same between all these formats. You know that. We've had read and had similar dialogues like this before.

However, I'll add something something different this time that I've noticed about those numbers on the BDP-1 and something else by other manufacturers. I'm not sure if those CPU% values solely can be relied up as an indicator for the resulting SQ.

I'll give you two instances. When switching from MPD to Roon Ready, the process itself takes a few seconds and in Roon application on Mac the BDP-1 will show up and you can get audio coming right away. However, on BDP-1, the computer's CPU% will spike and be high for quite awhile. Eventually in a minute or two, it will come down. For me, it will come down to 14-15% (or whatever the number happens to be during that cycle) and be stable at that.

Do you notice a difference in sound in Roon in the first two minutes as the computer's CPU% is very high?

In another case, when Roon has been playing for a long time and it's all stable, and then all of a sudden you open up Manic Moose and go to the Services Tab, you might see the CPU spike for a second or two and then it comes down very quickly.

In both of these instances, the CPU can spike well into the double digits and often over 100%. Of course, the transition from MPD to Roon has a longer impact on the CPU% displayed.

Do you notice any difference in sound on your unit? I honestly don't. In these cases, the numbers are much higher than what's displayed with either FLAC or WAV. ALAC is something else (I'm not sure if anything's changed as of late), yet no difference to me. I'm not sure if these numbers are 100% representative of the story.

However, I can often hear differences in WAV and FLAC. I've even made blind playlists with shuffle feature on Manic Moose in the past.

For example, Resonessence has spent quite a bit time working on the FLAC vs. WAV with their FPGA. I think it was initially all WAVs. There is definitely something there. I've seen similar discussion regarding Antipodes (including literature) and countless on Naim with some good explanations that go beyond basic CPU numbers and I/O.

Ultimately, I don't know how much of this matters to the average audiophile. Some care about the sound only and don't care about figuring out why something is behaving as it is or if it's not behaving as you would expect to. Similarly, others only go off of theory alone and think everybody is hallucinating. I like listening and figuring out what's what.

FWIW, I'll never recommending ripping CDs to WAV. It has to be FLAC (compressed or uncompressed) for metadata purposes. You can always derive a second WAV library if needed.

I think the processing of FLAC and WAV is different in MPD in comparison to Audirvana Plus. I remember Audirvana tries to decode everything as quickly as possible and buffer as much in advance. Is MPD decoding periodically in chunks?

For each device and protocol, it's good to know whether they buffer in chunks for quick responsiveness or buffer as much as possible in advance (which can have spikes initially).

CanadianMaestro

  • Restricted
  • Posts: 1458
  • Skepticism is the engine of progress
    • Hearing Everything That Nothing Can Measure
Re: BDP-2 and 3 and Roon
« Reply #21 on: 3 Apr 2018, 09:37 pm »
This is rationalization, not scientific analytical objectivity.
Still, you deserve an Award for Effort.  :thumb:

Krutsch

  • Full Member
  • Posts: 546
Re: BDP-2 and 3 and Roon
« Reply #22 on: 10 Apr 2018, 01:49 am »

However, I can often hear differences in WAV and FLAC. I've even made blind playlists with shuffle feature on Manic Moose in the past.


If you can run a blind test w/ WAV and FLAC and reliably hear the difference then you are on to something. I would love to see proof that such a difference is audible.

But the objectivist in me knows this cannot be the case. FLAC is a robust, lossless audio codec that has been tested to death. The PCM that comes from FLAC is EXACTLY the same as the PCM stored within a WAV container. Period.

The comparison of CPU utilization during playback is trivial between the formats. I personally don't believe this can result in audible differences; either from jitter or from electrical artifacts (i.e. line noise). But, I can't say that I've run experiments that exclude the possibility. So, believe what you will.

Audiophiles are fascinated with the prospect of discovering the next tweak - lately, this seems to be improperly grounded power supplies for network switches and/or improperly shielded or unshielded (I can't recall which) ethernet cables.

I spent many years in the Enterprise Storage space (working for a SAN vendor) and know that even highly dense switch configurations can manage gigabits per second of data, PER PORT, using cables that aren't even of BlueJeans cable quality. We tracked data errors and re-transmit rates, et al. All I can do is shake my head at an audio community that worries about artifacts from a single port running, what? 10 megabits per second of data?

EDIT: fixed typo on mega/giga BYTES to BITS.
« Last Edit: 14 Apr 2018, 01:40 pm by Krutsch »

CanadianMaestro

  • Restricted
  • Posts: 1458
  • Skepticism is the engine of progress
    • Hearing Everything That Nothing Can Measure
Re: BDP-2 and 3 and Roon
« Reply #23 on: 10 Apr 2018, 02:19 am »
zoom: If I drop an apple from my hand, I do not have to look (or do an experiment, blind or sighted) to know that it will drop to the ground. Physics. Inviolable.
Bits are identical between lossless formats like Flac and Wav. Physics. Inviolable. So they must sound identical (assuming all else remains basically unchanged).

But if you gain self-satisfaction in doing the expts (a bit like the satisfaction that comes from the RITUALS of playing vinyl), and hearing a "difference", then do it. No harm done, I guess, as long as your findings are not stated like Gospel.

cheers

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #24 on: 10 Apr 2018, 04:16 am »
I don't know why I always get a response telling me that bits are bits or the PCM streams are identical, when I always state that to be the case from the start. No matter how many times I state it and other things, it all seems to be ignored. :roll:

Here's what I'll respond with. Going with the logic from above:

- DLNA, MPD, Roon should all the same as they are all bit identical and result in the same PCM stream. Do they? Do you all agree that they sound the same? They should sound identical according to the logic...yet I still see many people here saying "MPD playback with flash drives" or "MPD sounds better than Roon"...explain that to me.

- Using a BDP to get better SQ should not make sense in comparison to using a standard computer for playback. You should be able to use a Foobar or VLC on Windows and Mac and get the same sound in quality. They both transmit bits perfectly. The DAC should be immune right? So why is everyone here spending so much on a BDP? Does a normal computer sound the same as BDP playing the same files? Everyone on other forums seems to tell me that they should sound identical. Do you guys agree that they all sound the same?

Everyone's free to use whatever they want or try or not try anything. I personally like trying experiments and often do blind tests and do take part in audio tests made by credited mastering engineers. I'll often give feedback for blind audio samples with wire testing or clocking. Having good ears and equipment is definitely a part in this. However, when it comes to the act of blind testing, you also have to get comfortable and need practice to perform well in this environment and have a strategy on how to focus or not focus (just focus enough). There's a learning curve.

Speaking of jitter, we often discuss whether jitter exists or not and whether it is audible or not. How about doing tests where you know in advance there is jitter present in the file. Can you figure out which file has lower or higher jitter?

http://www.cranesong.com/jitter_1.html

I've had only 1 person (a mastering engineer) so far report back to me. Almost nobody on any of the hi-fi audio forums seem to bother actually doing any tests and reporting back. The pro community is much different and better at this. I wonder why  :scratch:

FWIW, I use FLAC with Roon. I haven't changed the setup at in the past weeks or felt the need to tweak with this particular chain. YMMV. Everyone's free to choose. :thumb:

CanadianMaestro

  • Restricted
  • Posts: 1458
  • Skepticism is the engine of progress
    • Hearing Everything That Nothing Can Measure
Re: BDP-2 and 3 and Roon
« Reply #25 on: 10 Apr 2018, 10:24 am »


I've had only 1 person (a mastering engineer) so far report back to me. Almost nobody on any of the hi-fi audio forums seem to bother actually doing any tests and reporting back. The pro community is much different and better at this. I wonder why  :scratch:

1. They lead busy lives.
2. They'd rather enjoy the musik and relax than deal with "blind" testing, swapping wires etc.

The pros are different -- they make a living designing gear and selling it. It's in their interests to "test" stuff.

cheers

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #26 on: 10 Apr 2018, 05:08 pm »
1. They lead busy lives.
2. They'd rather enjoy the musik and relax than deal with "blind" testing, swapping wires etc.

The pros are different -- they make a living designing gear and selling it. It's in their interests to "test" stuff.

cheers

Right, you'd think so! Yet these same people find more than enough time to continually respond day in and day out. When it comes to taking a test that would take 5 minutes at most, silence. *Crickets*  :lol: :lol: :lol: :lol:

Yes, for the most part pros are different, thankfully! You can actually discuss substance with credited mixing/mastering engineers that have trained ears and listen to actual stuff...something that seems to elude most hi-fi forums.

CanadianMaestro

  • Restricted
  • Posts: 1458
  • Skepticism is the engine of progress
    • Hearing Everything That Nothing Can Measure
Re: BDP-2 and 3 and Roon
« Reply #27 on: 10 Apr 2018, 07:09 pm »
These "tests" take a lot more time than responding on a forum like this one.

99.5% of the human population cannot discern jitter that's less than 100 ps in a real-world home system (or any, in fact). That's a fact. So, if you're in that 0.5%, .....congratulations. Barnum and Bailey would love to hire you.
Cheers

Next.

Krutsch

  • Full Member
  • Posts: 546
Re: BDP-2 and 3 and Roon
« Reply #28 on: 14 Apr 2018, 01:44 pm »
I don't know why I always get a response telling me that bits are bits or the PCM streams are identical, when I always state that to be the case from the start. No matter how many times I state it and other things, it all seems to be ignored. :roll:

...


Zoom, if you re-read my response, I am questioning the statement that WAV and FLAC provide playback w/ audible differences. Note the portion about CPU utilization, et al.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #29 on: 14 Apr 2018, 05:32 pm »
Zoom, if you re-read my response, I am questioning the statement that WAV and FLAC provide playback w/ audible differences. Note the portion about CPU utilization, et al.

That's why I said doing the blind shuffle test in Manic Moose. I've written about this even two years back. Pick a couple of songs and have them converted to FLAC and WAV via XLD. Put them on the same flash drive and folder. Manually add each Track A.flac 10 times and Track B.WAV 10 times. Put it on shuffle. Press play or use a Harmony remote and try to figure out which is FLAC or WAV. You can get creative and try to setup the Manic Moose window in a way that you can access the essential functions while hiding the other details.

I even wrote at that time that it can be very tricky and that on certain tracks, I repeatedly guessed the FLAC as the WAV. Why? If the assumption is that WAV is the better sounding one, that doesn't always hold up, as with certain material the "character" of FLAC can actually work in favour of the track, especially in quick AB testing.

You also want to pick songs that have valuable information in the first 10-15 seconds that can be useful in picking one from the other. Ambiance is a good thing to focus on. Time intervals with this sort of stuff is absolutely critical. It's easy to fatigue the brain and make it go numb. This is why it needs to be done on multiple days.

The rig does need to be resolving. Sometimes, even if the components are good, simple things like digital cables and length can compromise and mask the chain. I've had a talk with Grimm about this and they've found similar results. I think most people assume that their gear isn't susceptible as it sounds good and work correctly, but they probably haven't bothered to test that out. They might be very surprised with some of the results and what they are missing out on. Also, headphones can sometimes be good for certain tracks, while on others they mask other stuff that speakers provide or present it in a way that doesn't fully connect with me.

I haven't done this level of testing since Roon. As for the CPU%, I'm not sure if those numbers in the Services Tab are 100% representative as indicators of SQ. As I mentioned, simply opening the Manic Moose page and going to services tab can spike the CPU% well into +70%. Don't really notice a difference there. Same thing when switching between Roon and MPD in the first 2 minutes where the CPU%. You'd expect the sound to improve over time, but it's not there.

I'm still not 100% sure how (much/often?) the buffering is done with MPD, in comparison to something like A+.

If you had an album of say 10 tracks and loaded them all and just pressed play once at the start and letting the album end without any interruptions, how would the overall processing be different in Audirvana Plus and MPD? I've had designers tell me that the higher I/O with the WAV is nowhere near as impactful as dealing with FLAC when woking on their FPGA. At the end of the day, it comes down to whether it's audible or not. It can be done both sighted and blind, if one cares.

These were the questions I was asking but never got a response to when I was still on MPD. Thankfully, I don't have to worry about formats anymore with Roon.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #30 on: 18 Apr 2018, 08:19 pm »
Soundstage and Plane

With this whole Roon and MPD, and ethernet cables, and FLAC vs. WAV etc... it can be hard to quantify subjective things and make sense of what we are hearing. Is a change for the better or worse? I suppose it's ultimately up to the listener's preference.

I'm still not sure what parameters are associated with what, or whether one has less or more distortion, or what mechanism is at play with each other. I've speculated before, so I won't do more of that...I'll just state my observations.

With typical compressed music, it can be hard to make sense of things and you sometimes can flip flop back and forth depending on the recording. Recently, I spent a lot of time with large orchestral pieces and stuff with loads of dynamics and sense of space. Lots of John Williams as well and good vocal Jazz. High dynamic range material.

My observations were very similar to what was said here by James Tanner in comparison to BP-17 and BP-26:

Quote
I find there are a few differences between the BP26 and the new BP-17 Cubed preamplifiers.  The BP-17 Cubed circuitry is based on the new patent we received for the input stage of the new Cubed amplifiers.

The soundstage presentation is different – with the BP-26 it’s a more up-front sound with the voices or instrument tending to be at or slightly forward of the plane of the loudspeaker whereas the BP-17 Cubed has a more recessed presentation where most of the action is at the plane of the speaker and further behind.  The size of the stage  (left to right and front to back) as well as the presentation of the instruments (size and position) within the stage was very close so no major difference there I could detect. The BP-26 sounds a little ‘puncher’ in the mid-bass frequency area while the BP-17 Cubed is a little less forceful. If you like that ‘in your gut’ bass punch you may prefer the BP-26 in this area.

The part where I think the 17 excels is when a crescendo comes along – either voice of instrument. In some recordings where the vocalist or instrument goes from a low or medium volume level to a crescendo or maybe even approaching a scream with voice  the 17 never sounds strained or has any sense of glare or stridency. Sometimes I was thinking the BP-17 Cubed lacked dynamic capability but the more I listened the more I thought – no – it’s just that it is not producing any sense of overload or strain in the music when pushed. Transients in real music are huge and the ability of the BP-17 to deal with these huge shifts of level is something it handles very well.

So if I had to sum up the BP-17 in a word it would be “organic” – less hi-fi sounding. It has a very natural sound to it with no strain or glare.

Taken from here: http://www.audiocircle.com/index.php?topic=151889.40

With large orchestral pieces and good recordings, it finally became very clear what I had been feeling was happening with some less than stellar music. If you listen for more details vs. less details, or more bass vs. less, or tighter vs. less bass...your thinking and results will differ depending on which criteria you are using in your head for assessment.

I find that between the FLAC vs. WAV or long vs. short ethernet cables, the of plane of the music can change. In some instances, when the plane recedes, the music can initially feel as if it's lost some impact or if some details are lost. In quick AB testing, you might very well prefer the presentation of the more up front music as it comes across as louder or more detailed. However, as you listen more and more to each, the one that initially seems less detailed ends up being the more easier to listen to and enjoy and just flows like liquid. On the other hand, the more exciting one can start to sound harsh or tiring over time and even constricted.

It'd be interesting to have measurements relating to how we perceive the soundstage or the plane of the soundstage.

I've tried the same list of tracks in both MPD (with FLAC and WAV) and Roon with different setups and cable lengths. On top of that, I tried using my iMac as a source. I used the same music with Audirvana Plus. Then I also found Spotify and Youtube versions of those tracks which were more compressed and lossy and being fed by a worse transport (iMac).

The variance in how the plane varied between all these combination was quite surprising. In general, whenever I use my computer as a source and listen to music or watch movies or clips on Youtube/Netflix, the plane of sound is always in front of the speakers. The iMac as a source is always the most up front.

With FLAC and WAV, I find the FLAC is more up front. With short ethernet cables, I also find the soundstage more up front.

Right now with this particular Fast ethernet switch (DES-1005) powered by Teradak LPS and long ethernet cables (30+ feet) for each of the links, I find it's the easiest to listen to and the most fluid of the bunch, and with a soundstage that is behind the speakers. If you are initially switching from a more upfront presentation, this can initially sound flat and less dynamic. This is similar to what I also notice with my DAC and BDP-1 when swapping out equivalent digital cables (same wire + connectors) for short and long ones (2 feet vs. 6 feet vs. 10 feet vs 18 feet for SPDIF and AES).

If you are listening on headphones exclusively, good luck! This is something that doesn't quite make sense and translate on headphones.

So the next time anyone wants to do a comparison, maybe they should try and focus on the plane of the music as well aside from tonality. Orchestral soundtracks are great for this.

I'd be curious to see measurements on distortion and other stuff vs. where/how one perceives the plane and soundstage while using the same material and controlling for other things.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #31 on: 19 Apr 2018, 07:58 pm »
Here are some articles from Paul McGowan about soundstage and where the image should be and the logic behind it:

https://www.psaudio.com/pauls-posts/music-supposed/2495/

https://www.psaudio.com/pauls-posts/depth/7796/

https://www.psaudio.com/pauls-posts/depth-2/

https://www.psaudio.com/pauls-posts/agreeing-on-little/


I'll be honest, when I'm into the music I never think about soundstage. It's more about the overall flow or what works. It's subconscious. If you have to focus on it, you lose the sense of what you're trying to achieve. It's only during testing where I try to make sense of it as it relates to my preferences. When I'm listening and even comparing, I'll usually play one configuration and then not think about it while listening semi actively. I'll start reading some other stuff and just let the music do it's work subconsciously in the background. It leads to more reliable results for me over when I actively try to listen and focus on things. That goes completely against how we typically hear our surroundings in day-to-day activities. We are never listening for something actively. Instead, we just listen and react. Same thing goes for any activity. Take running for example. If you have to mentally tell yourself to put your right foot in front and then followed by left and then right and so on, your performance will go down fast. You in fact don't want to think. It's more a trained response that comes with practice where you can do it better and better without thinking.

Anyways, enough about that. As far as these tests are concerned, I'll mention one other thing to consider when evaluating different methods: SPL. What material are you listening to and how loud are you listening?

Play something quiet, and sometimes the added distortion or something extra can sound more detailed simply because there is more to it. However, when you start turning things up on a full range system in a big room, that's when things start reversing. That extra noise/distortion will turn into strain and sound restricted in comparison, whereas the cleaner thing will sound easy to listen to and without any strain with a sense of higher dynamic range.

:thumb:

Krutsch

  • Full Member
  • Posts: 546
Re: BDP-2 and 3 and Roon
« Reply #32 on: 20 Apr 2018, 03:23 pm »
...
I haven't done this level of testing since Roon. As for the CPU%, I'm not sure if those numbers in the Services Tab are 100% representative as indicators of SQ. As I mentioned, simply opening the Manic Moose page and going to services tab can spike the CPU% well into +70%. Don't really notice a difference there. Same thing when switching between Roon and MPD in the first 2 minutes where the CPU%. You'd expect the sound to improve over time, but it's not there.

...

For what it's worth, I am not using the services tab; I am ssh'ing into the BDP and using 'ps' and 'top' to look at CPU. It's not the best way to study CPU utilization, but it's representative of the utilization differences between track playback.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #33 on: 6 May 2018, 07:45 pm »
Ok, quick update first. I've been back on stock power cables for some time now. Got rid of the shielded cables which I found problematic. Also, the BDP-1's been running without problems on the Torus for a few days without a single error code. The Torus isn't loaded any differently than before. The only difference I can think of is the longer ethernet cable that I am using compared to before (6 feet vs. 50+ feet). I'm guessing the longer cable consumes more power and so the BDP-1 can better displace any occasional spikes from the Torus. Anyways, everything plugged into the Torus and no ground loops.

I had to revisit some of my older notes and methodology. I think I have a better understanding now of why people dislike Roon's sound quality or playback NAS (both through ethernet) and prefer direct connected drives, or why Karl Schuster prefers FLAC over WAV/AIFF and hard drives over flash drives or SSDs. He finds his respective picks more musical than the other. He felt the same distaste about pulling music via ethernet either for DLNA or MPD. The other methods in comparison sound more drier and cleaner but ultimately not engaging. While FLACs off portable spinning hard drive sound best. I might have to agree with that.

I think Bryston did an exceptional job with controlling noise as typically concerned and with power supplies and separating different boards and power supplies. This is why my observations regarding a few things have never changed regardless of what the setup was. I'll share them from both a Roon and MPD perspective. They both support each other and the overall isolation achieved by Bryston. You can also test this at home and if you find any differences, please do share!

Say you have MPD running say off flash drives. If you plug in an ethernet cable and start using it for interface control, it makes no difference to sound. The extra power consumption of the ethernet port makes no difference. Similarly, you can also plug in a HDD to another port and it will spin during booting. You'd think there would be a degradation because of this, yet nothing. You can do this the other way as well. Play off a hard drive and it doesn't matter what other drives are plugged in or what's happening with ethernet.

Even with Roon, you can put in any hard drive and have it boot up and it makes no difference. As I said before, ethernet cables and power supplies or switch used makes no difference with MPD if playing off USB drives. However, when used for NAS or Roon, all these things start making a difference.

How could that be? We'd assume that it was a "sum of all parts" situation, but clearly that isn't the case due to how well Bryston made the BDP. Other peripheral devices make no difference. Based on the observations, it's only the chain directly involved that makes a difference.

I've done tests regarding conducted and radiated noise already in another post. I can pile noisy stuff and 100 feet of ethernet on top of signal wires and it makes no difference. Similarly, pulling ethernet cables makes no difference when USB drives are being used.

Sidestepping for a moment, so why might FLAC or WAV sound different given that the load isn't much different for the CPU, especially when considering that pulling an ethernet cables makes no difference. Surely, the power consumption difference FLAC/WAV is nowhere near powering the ethernet port. On the other hand, why might some people like FLAC over WAV. Or why might people like portable spinning hard drives over flash drives. You'd expect hard drives to be noisier and consume more power. Why might one flash drive sound different than an other flash drive? Similarly, I could use the same power supply for the switch and same cables, but swapping out a gigabit for a fast ethernet might result in different sound. Or why the same cable in different lengths can sound different.

However, as we've noticed that the sum of the parts argument doesn't hold up due to the implementation of Bryston regarding noise.

I am not capable from an engineering standpoint to convey or explain what I think might be going on. Although based on the testing above for noise in both USB and ethernet realm and ruling things out systematically, I think the key might be jitter here (what domain or where to look? Not sure). I've read comments from John Swenson about this and he'd be the person if you wanted it answered any better. It very well could be the oscillators in various ethernet devices and USB flash drives that are imparting a particular AND constant pattern (of noise or jitter? I have no idea what to label it as) that may be going all the way through the chain and not being considered.

This might explain why FLAC might be preferable to WAV. WAV has a constant and periodic pattern. FLAC on the other hand has a random pattern and is distributed all over the place. Having that jitter spread all over the place might make it less or completely inaudible. On the other hand, a constant and specific pattern with WAV might be more noticeable. Similarly, flash drives may be suffering from this as well due to the oscillator. I'm not sure about hard drives in this regard, but 2 years ago when I knew nothing about any of this in my testing I actually preferred the hard drive as more fluid and musical, but acknowledging that the flash drive sounded 'cleaner.' There is a 'convincing' aspect to this that Karl Schuster also wrote about that I have to agree with. Spinning drives and FLAC.

SSD and USB Wifi adapters also have oscillators that vary from one another. Just thinking out loud. Maybe this should be looked.

I wonder what Bryston has to say about this? Do they maintain that all these devices and listening methods sound the same? I think James has given his preference before about MPD over Roon, and USB flash drives over hard drives. So in some cases perhaps? I'd love to get feedback from Bryston or anyone. It would be good to have confirmation by others as to whether they also get the same immunity by plugging and unplugging other things that are not being used in the signal chain.

I'm not sure if these things have been looked at based on the reviews I've read so far for Bryston BDP-3 even when used with BDA-3 DAC together, which is the best so far in terms of jitter rejection. This could be a good thing to look at for BD(A/P)-4.

R. Daneel

  • Full Member
  • Posts: 773
Re: BDP-2 and 3 and Roon
« Reply #34 on: 12 May 2018, 07:44 am »
Jitter couldn't possibly have anything to do with that. Data stored on memory chips and hard drives is void of timing information. The data is timed correctly once it has been processed by the audio circuitry located on the daughter-board of the BDP - if we are talking about using BDP's AES/BNC outputs / or in the USB processor within the DAC itself - if we are talking about using BDP's USB output.

According to Audio Engineering Society (AES) whitepaper jitter cannot have any effect on audio if it is below 0.5 nanosecond (500 picosecond). Modern-day DAC's have a jitter numbers well below that which just tells me how obsessed with numbers this industry is. A solidly engineered input receiver section within a DAC will eliminate almost all jitter regardless of signal quality it has been fed.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #35 on: 12 May 2018, 08:37 pm »
I'll admit two things:

1) My DAC may not be the best, especially compared to what's available right now in terms of jitter rejection.

2) I still don't know why all these methods sound different from one another. From a practical standpoint, one can decide what's best for them by listening. If someone wants to find out for themselves without measurements, they can setup logical experiments to try and better understand what may or may not be happening.

Having said that, while my DAC may not be the best, all the professional reviews and user comments that have done critical comparisons have found similar differences. This is even the case with the latest BDP-3 and BDA-3 DAC that should be even more jitter/noise immune than past DACs and digital players, yet they are still there. Are all these people wrong and delusional? Without measurements and knowing where to look, it's impossible to correctly address what might be happening.

Recently, Lumin released their new player with fiber inputs to address this.

Going back to practical matters at hand: I know I prefer Roon's interface over MPD and so I try to make that sound that best. There's two (actually three) ways of doing Roon on the BDPs.

1) Ethernet
2) Wifi USB Dongle
3) Built in Wifi (BDP-3 only)

Right now, I am giving the USB Wifi adapter a shot again. In the past, those shielded power cables might have skewed the results. No routers, switches, ethernet cables, USB drives, clutter - just a simple USB stick attached to the BDP-1. I'm liking it so far. It's been stable with Roon and having everything on the 5 Ghz network.

Krutsch

  • Full Member
  • Posts: 546
Re: BDP-2 and 3 and Roon
« Reply #36 on: 13 May 2018, 03:13 am »

...

Going back to practical matters at hand: I know I prefer Roon's interface over MPD and so I try to make that sound that best. There's two (actually three) ways of doing Roon on the BDPs.

1) Ethernet
2) Wifi USB Dongle
3) Built in Wifi (BDP-3 only)

Right now, I am giving the USB Wifi adapter a shot again. In the past, those shielded power cables might have skewed the results. No routers, switches, ethernet cables, USB drives, clutter - just a simple USB stick attached to the BDP-1. I'm liking it so far. It's been stable with Roon and having everything on the 5 Ghz network.

One thing to keep in mind: Ethernet is full-duplex while Wi-Fi is half-duplex. This means that bi-directional traffic happens simultaneously with Ethernet; with Wi-Fi, traffic only moves inbound or outbound, at any one time. For a "busy" protocol like RAAT (Roon's protocol), that may be more important than measured throughput in any one direction.

By "busy", RAAT has a fair amount of back-traffic from the end-point to the Roon core / server (updating the UI, requesting packets or timing information for multi-room play, who knows?).

And, of course, Wi-Fi reported signal strength and connection speed don't usually equate to actual performance.

I am not saying you can't enjoy Roon via Wi-Fi - you can - but be aware of above comments when deciding on how to connect your Roon Core/Server and your BDP-1.

zoom25

  • Full Member
  • Posts: 770
Re: BDP-2 and 3 and Roon
« Reply #37 on: 13 May 2018, 04:31 am »
One thing to keep in mind: Ethernet is full-duplex while Wi-Fi is half-duplex. This means that bi-directional traffic happens simultaneously with Ethernet; with Wi-Fi, traffic only moves inbound or outbound, at any one time. For a "busy" protocol like RAAT (Roon's protocol), that may be more important than measured throughput in any one direction.

By "busy", RAAT has a fair amount of back-traffic from the end-point to the Roon core / server (updating the UI, requesting packets or timing information for multi-room play, who knows?).

And, of course, Wi-Fi reported signal strength and connection speed don't usually equate to actual performance.

I am not saying you can't enjoy Roon via Wi-Fi - you can - but be aware of above comments when deciding on how to connect your Roon Core/Server and your BDP-1.

I have zero problems with Wifi here at my home. I've used it already for weeks. The experience is identical between ethernet and wifi at my place. I can skip between tracks and different albums and jump 3 minutes into the track right when I skip into the track and it's instantaneous. I have tried to go out of my way to make it fail on my end and it's never happened. Zero dropouts. I have both my iMac and BDP-1 connected to the network wirelessly.

For me, the only question that remains is sound quality. I compared a few tracks from Little Dragon last night with Roon via my Wifi adapter and MPD with flash drive and WAVs. I tried to give each method the best possible shot. Nothing else plugged in aside from the respective USB sticks. No ethernet cable or other storage drives plugged in. MPD has everything disabled. As bare bones as possible for best SQ. At the end, I ran both MPD and Roon simultaneously to see how they would compare in even faster switching or whether it would mess with the SQ due to the interaction of both flash drives and wifi adapter plugged in at the same time. I gave it a thorough examination.

I also briefly tried the USB Wifi adapter with Jitterbug in the middle and did not like it at all. I'm plugging it straight into one of the front USB ports.

I'm pleased with it so far. Haven't felt the need to break out the whole ethernet setup of cables switches and LPS so far.

For those that may be interested, this is the adapter I'm using: http://a.co/dqQlimU

"Panda Wireless Pau07"

TJ-Sully

  • Full Member
  • Posts: 355
Re: BDP-2 and 3 and Roon
« Reply #38 on: 13 Mar 2019, 10:44 am »
hi folks,

found this older thread.

after all this in depth discussion last year, has anything changed regarding the best way to configure Roon, either physically or through software?

I use ROON almost exclusively for my listening, and I know the SQ can be better.

thoughts folks?

T