Why Do Some Recordings Sound Great and Others Dead?

I listen to Radio Paradise MQA on my NODE 3, SMSL M400, B and O Beolab 8000’s and Hsu 15in sealed sub.  The acoustics in my room are poor.  I’ve noticed that some recordings sound very realistic. For instance the vocals on a Stabat Mater dolorosa hymn sounded great.  But a Nora Jones recording was terrible.  Her voice was lost in back of some murky instruments.  I’m familiar with this recording listening to it on my iPod, where her voice shines out and the music is good.  I’m wondering why the big differences?


I listen to Radio Paradise all the time… through Qobuz for a couple hours a day. My streamer / digital system sound the same as my really great analog end (you can see my system under mt UserID). No, I do not experience any significant difference in sound quality among their play list. All sound quite spectacular. 

Recordings vary widely in SQ, as I'm sure you know.  Some seem to agree with your system and room while others don't, and some will never sound good.  They're just bad recordings.

You say the acoustics in your room are poor.  Can you do anything to improve the room?  That's probably what you're going to have to do.

I listen to RP a lot and the quality is pretty consistent, especially on the Mellow Mix station. Question, how are you controlling volume in your chain? Node, SMSL, Beolabs?

I believe the only way I can control the volume is through the SMSL DAC.  I use the SMSL remote.  I bypass the NODE DAC and use a USB cable to send the signal to the SMSL.  Under ‘Audio Settings’ and ‘Select Output’, once I have connected the NODE to the SMSL it reads ‘Default’ ‘SMSL USB AUDIO.’  The Beolabs have no volume control.  For my subs, they have their own volume settings for their volume in the system.

Too much compression (dynamic range), overuse of limiters, microphones not set up properly, bad mixing engineer, bad recording engineer.

Or it could be the (most likely) Wolfson DAC chip in your iPod. Many people agree that Wolfson DAC chips are on the sweeter side of neutral, rather than clinical sounding.

[Verse 1]
Sunrise, sunrise
Looks like mornin' in your eyes
But the clock's held 9:15 for hours
Sunrise, sunrise
Couldn't tempt us if it tried
'Cause the afternoon's already come and gone

[Chorus 1]
And I said
To you... 

great song

A couple suggestions:

1) I assume you have 'Audio Level Fixed' set on the Node? Not sure that is settable if you bypass the DAC, but make sure it is set to pass through.

2) Try listening to RP's FLAC lossless stream instead of MQA. MQA adds a lot of complexity, maybe some songs are not being expanded correctly.

3) Looking at the M400 manual, it is pretty lean on explanations, but I see reference to various filters, phase and brightness settings. Not sure these are all relevant, but maybe some adjustment could make a difference.


Revision, I see that phase applies to balanced and brightness is for the display. But maybe the PCM filters are relevant.


@ghdprentice Thx for your comments, as it’s good to know that you have consistent excellent sound quality. My system is basic, but I’m tinkering to see if I can fix a week spot. I just ordered parts to create a fiber optic link to my NODE.

@tomcy6 I noticed that older recordings from around the 60’s often don’t sound that well. I’m surprised at the Norah Jones recording, as it’s more recent and I am familiar with it. I don’t think that it’s the recording - it’s just not coming to me well from Radio Paradise. I just added a thick rug to my room, but it’s mostly glass, so it’s never going to have good acoustics.

@zlone Yes, I’m on ‘Audio Level Fixed.’ I tried the SMSL filters and really don’t hear any difference. I’ll try the FLAC stream.

@mastering92 Thx for singing! Norah sounds good on the iPod. But why would one vocal recording sound good on my system, but not another that I know is a good recording from another source?

I will probably leave out from my Favorites list any recordings that don’t sound good, like Nora’s. I have found many others that sound excellent, so I’ll still have plenty to enjoy.


Which Nora Jones song was it? I could check my redbook audio collection...cd rips.

Now I want to listen...and maybe I’ll have an answer to your question.

@mastering92  ‘Light As a Feather’.  I have it on CD and my iPod, where I’ve listened to it every other day or so for the last 10? years.  On Radio Paradise her voice does not sound very ‘present’, but is lost and muffled, as are the instruments.  On this one, I’m not getting what I expect from past listening.

In ’I Put a Spell On You’ Samantha Fish’s voice shines for the most part, but there are a few portions of the song where the volume recedes so much it sounds like the system is going out.

On the other hand ‘Lullaby’ by Pieta Brown sounds like a very good overall recording, her voice stands out from the instruments, and both her voice and the instruments sound great.

 I tend to listen to music as lower levels.

Recordings are mixed per the studio's speakers. They might agree with yours, they might not.

what @fuzztone said

I don’t know about the other 2, but "light as a feather" is a great song and I’ve heard it many times before.

Her vocals are clear and forward. Her voice peaks quite a bit, but her voice is balanced. Instrumentals take a backseat The track moves at a slow pace, like a light feather falling! Sound character overall is slightly warm. Not too bright.

I agree it’s a great song, so it’s either Radio Paradise or my system that is the weak point.

’’When I Was Older’ by Billie Eilish is a delicate female vocal that sounds great on what I have.  I’ll keep exploring solutions and/or skip songs that don’t sound well.

It’s a common symptom of narrow, but tall peaks in the mid-bass or bass. Clipping them with an EQ/DSP is often the answer.

The reason it affects some recordings but not others is because the peaks are pretty narrow, so many recordings don't trigger the bass node.

While you CAN fix this with a variety of ways, like moving a sub, speakers, bass traps and moving your listening location, these particular symptoms are easy to deal with via DSP. 

Of course, other methods may give overall a better, broader solution.

One more thought, as an experiment you might try removing your DAC from the chain and let the Node do everything to see if you still hear significant differences between songs. 

@erik_squires  Thanks for your suggestion!  I’m not surprised that my room has difficult ’issues’. I would have thought that they would manifest themselves with all recordings rather than select ones.  I’m not very experienced in audio, and not very good at A:B comparisons.  I have 2 sealed subs, a 15in Hsu and a 12in Paradigm.  Generally I think my sub-bass is excellent.  I have them turned down so I don’t have too much of it.  I looked a a minidsp ddrc24, and was put off by what appears to be a very complex process to implement it.  Maybe I’ll look into that or something with Dirac.

I have them turned down so I don’t have too much of it.

This is exactly part of the syndrome!! 🤣

When you clip the peaks, the subs will sound too low and you will want to bring them up in level but now they’ll sound super smooth and natural.

This article I wrote will help you.  It is more about fixing bass problems than avoiding a sub, so if you read through the tech parts you'll know what to do.




For what its worth, I use an Analog Production vinyl pressing of Nora Jones' 1st lp to show off my system to friends. Sounds great! So the original recording is very good. Not sure if this helps as a reference for streaming this recording.

i've noticed there is a wide variety in sound quality from recording to recording, even those made in the same studio with the same people. a lot can go wrong between the studio and your listening room, at multiple steps along the way. some recording studios [particularly classical labels] try to ameliorate this situation by using consumer equipment in a consumer-type of studio listening room, to make their sound something that sounds more reasonable in a typical home listening environment. i have found that for most situations, a certain modest amount of dynamic range compression/limiting is helpful to make a given recording sound the best in the universe of different listening environments/equipment complements out there. there is an artful way of doing this which doesn't do [as much] dirt to the lucky owners of megabuck equipment and megabuck listening rooms. a light touch is required, but not all recording professionals have a light touch, as witnessed by the thousand and one sins one hears in commercial recordings. the biggest sin IMHO is mixing the vocals in a way that they are barely discernible from the musical background, i believe the germ of this idea came from hollywood movies with their emphasis on "realism" IOW maximizing the [relative] impact of the loud parts by reducing the volume/clarity of dialogue. a lot of artists don't care to have their vocals too high in the mix [Elvis was a notable example], and the engineers are all too happy to play along. i have [per audiological exam/spectrogram] normal hearing for my age, so when i hear muffled vocals listening on my Sennheiser HD580 cans, i know something is wrong with the recording. 

@erik_squires  Thx for your explanation.  I reread it, and I can appreciate the considerable time and expertise in your analysis.  I don’t currently have enough knowledge to follow up on it with any confidence. My room is a mess, but I  can’t do much to improve it.  I tried following various suggestions and acoustic placement principles, but with 2 Beolab8000 speakers, 2 subs and an unwieldy room, I’m just guessing.  I’ve worked on speaker microadjustments,  Next I’ll try the subs.  Since my speakers and subs are all powered, I’m not using a preamp with the SMSL M400.  I wonder if that could be a problem? Maybe some day I’ll have more confidence to try minidsp or Dirac.

@emrofsemanon  Thx for your comments!  I just listened to May Erlewine’s ‘Days Gone By’ and her voice was very forward, loud and clear.  I believe the same is true of the Nora Jones recording mentioned above - but it just does not come through on my stereo system.  Until I get it ‘fixed’ I’ll just listen to what sounds best for me. 

@OP - If you feel uncomfortable, here's my recommendation:


Corner Bass traps or soffit traps from GIK.   The soffit traps work a little better, but the corner bass traps give you more floor space.

Not sure if RP is the a good reference or not as I don't use that service, but there are a couple things to consider. 

- If you don't know the CAT # of the specific album, well that is a major prob.  That is also a big downside to streaming - you have no idea (many times) of the provenance as well as they can remove/change said album version at any given point and you'll never know (generally speaking)

- The recording/mixing/mastering means everything (provenance) and if one doesn't know that then all bets are off

- Some recordings most def sound better than others (I would hope so)?  However, as has been pointed out, the room/acoustics are paramount to SQ along with proper setup.  Those two items are an absolute MUST before anything else. 

i also believe in "tone control courage" IOW using definitive electronic means to correct an incorrect frequency balance in a given recording. 

With my current setup I don’t have any tone controls.  The NODE has some when using the internal DAC.  I’m bypassing that into the SMSL DAC, which has no tone controls.  I don’t have a preamp, and my speakers are powered.  I’d have to put something into the system like a Lokius, minidsp or minidsp ddrc24.  There are some crossover frequency and q controls on my subs.

Some recordings sound great and others dead because of differences in recording, mixing and mastering.

I hear five different classes of sound quality in my collection of albums, even ranking each album D, C, B, A, A+ based on sound quality only.

This has nothing at all to do with the "music quality" of the material or the artists.

That's a great idea tom. I've got alot of used classical records from my father and an estate sale.  Ranking them sounds like a wise idea!

How do you rank them? Do you use post it notes or some other method?

@jjbeason14 I have found odds are good a classical or jazz record will be in class A, with some very spectacular recordings getting into class A+.

Most pop and rock are in class B, with quite a few in class C.

The only class D albums I have were recorded  before 1950.

Based on the responses, i don’t think recording quality is the reason for the differences I hear.  My system and room just don’t go well with some recordings.  Cupping my ears towards the speakers seems to improve almost all of them, but while some show just a slight improvement, others have a night and day difference, with a very significant, holographic improvement.