Madonna "sounds" harsh : due to recordings or my system ?

[tgb]

Hello,

You might not like her... I like her & her CDs (all), this is not my issue there ;-)

 

I find that my system sounds quite OK, I mean detailed but natural on voices & instruments. Fine for me.

But I am a bit annoyed with Madonna recordings, like MDNA for instance, which sounds rather harsh, and unpleasant during long listening.

Do you face the same trouble with her albums, or is it me or my system that are wrong ?

 

System :

data on NAS syno 212J with Mnimserver > sotm SMS100 > USB-spdif interface > Mutec MC3+ > Dangerous Source DAC > integrated amp Linn Majik-I > Kelinac 811MG ; SMS100 & DAC on Keces DC116

 

Thanks & happy new year

Rgds

[CatManDo]

But I am a bit annoyed with Madonna recordings, like MDNA for instance, which sounds rather harsh, and unpleasant during long listening.

 

That album suffers from high dynamic compression (DR6 in the dynamic range database), so it will sound harsh on any system:

 

Her classic albums from the 80's have also been compressed in the remastered CD reissues. The original CDs (and the HDtracks downloads) probably sound much better.

 

Album list - Dynamic Range Database

 

Welcome to the ugly world of the loudness war, and a happy new year

[Matty70]

I only have one Madonna album: Ray Of Light.

 

Despite the dynamic range it shows another significant problem which could be the reason for harshness too.

 

In the following MusicScope screenshot (Track 3: Ray Of Light) you can see massive occurences of intersample peaks and as a result potential distorsion almost all over the time.

You can see them in the historic circle, where the curve is red instead of green. Max. TPL is 0.9/0.6 (L/R).

 

The other tracks I checked look similar.

 

Regards

Matthias

 

[tgb]

THanks a lot Matty,

Great your software, I will investigate this too.

Rgds

[Matty70]

Hi tgb,

 

yes I really like this tool very much. It gives you a very good insight into many quality aspects of music material.

What I like most is the flexibility of usage, the compatibility to many file types, including DSD and that it is implemented with the EBU R128 and ITU-R BS.1770-3 standards.

 

I'm not involved in any way in the company or in the devolopment of this software. It has been developed by a friend of mine.

So I think I'm one of the first users and I am really enthusiatic in using it, especially to check my highres downloads.

 

Regards

Matthias

[Paul R]

I only have one Madonna album: Ray Of Light.

 

 

 

In the following MusicScope screenshot (Track 3: Ray Of Light) you can see massive occurences of intersample peaks and as a result potential distorsion almost all over the time.

You can see them in the historic circle, where the curve is red instead of green. Max. TPL is 0.9/0.6 (L/R).

 

The other tracks I checked look similar.

 

Regards

Matthias[/Quote]

 

That sure looks like a nice hunk of software. I just ordered it and will play with it tonight. I hate paying in Euros though, I always get charged a bit of a currency conversion fee.

 

-Paul

[89reksal]

Re: MusicScope

 

I also purchased it today. Thanks for mentioning it Matty70.

 

Bill Street

[rodrigaj]

Only in CA could we turn a thread about Madonna into a discussion on MusicScope. LOL

 

I purchased MusicScope last night and also got 10Euro coupon for any HIRESAUDIO download.

 

It is a very nice piece of software, but be aware that it does not analyze AIFF files.

[Paul R]

Only in CA could we turn a thread about Madonna into a discussion on MusicScope. LOL

 

I purchased MusicScope last night and also got 10Euro coupon for any HIRESAUDIO download.

 

It is a very nice piece of software, but be aware that it does not analyze AIFF files.

 

Yep, discovered that after I purchased it. The developers say they are working on AIFF support though. it's fairly useless to me without AIFF support, but nice when you can use it.

 

Looks like it has some trouble recognizing DSD128 too, it reports them as DSD64 here. At least on the few I tried last night.

[tgb]

Only in CA could we turn a thread about Madonna into a discussion on MusicScope. LOL

hi Rodrigaj,

you're right... if you don't like music ;-)

I like Bach and I consider Beethoven opus 111 as the greateast piece of music ever... and Madonna sounds great to my hear... ;-)

[cjf]

I own all of the HiRez Madonna recordings found on HD Tracks and think they sound delicious. I like her self titled first album the best in 24/192 format.

 

No large amounts of compression on these. I don't recall the DR values off hand but do remember them all being very respectable.

 

If you don't already own an HD Tracks version I would try one and see if they change your mind on the sound.

[copy_of_a]

Despite the dynamic range it shows another significant problem which could be the reason for harshness too.

 

In the following MusicScope screenshot (Track 3: Ray Of Light) you can see massive occurences of intersample peaks and as a result potential distorsion almost all over the time.

Allmost all 44.1/16bit files from Rock/Pop and similar Genres show values up to +1dB True Peak. Especially CDs from the 90th. Only very, very simplified and very low budget DACs will distort with True Peaks in this range. Some tracks may go up to +2dBTP or +3dBTP... and even higher (values above +1dBTP are typically NOT a result of heavy compression but of very fast changes in dynamics & attack in conjunction with max. digital values close to 0dBFS). A good DAC will handle this without issues... The Ray Of Light Album (the Warner CD from 1998) doesn't sound harsh on my system (it does sound compressed, of course, but not harsh). Also doesn't sound harsh on my simple Marantz CD5004 (which certainly doesn't sound bad ... but the DAC is of course not special in any way).

You can try to add a Limiter-Plugin that is able to limit only Intersample Peaks ... or simply try to lower the gain of the file by 1dB (in an Audio Editing Software or at least software-based in the player software with Dither applied) to see if this sounds better. If so, than your DAC may have issues with True Peaks. But I doubt +1dBTP will cause distortion on any mediocre DAC available today. I would bet the DAC of the OP is more than capable of handling these Ture Peaks...

[Skeptic]

Despite the dynamic range it shows another significant problem which could be the reason for harshness too.

 

In the following MusicScope screenshot (Track 3: Ray Of Light) you can see massive occurences of intersample peaks and as a result potential distorsion almost all over the time.

You can see them in the historic circle, where the curve is red instead of green. Max. TPL is 0.9/0.6 (L/R).

For what it's worth, inter-sample peaks are not really a problem as such.

You can fix them by simply reducing the playback volume in the digital domain.

Personally, I prefer to use JRiver which incorporates this analysis in its Volume Leveling engine, and prevents any inter-sample peaks clipping.

 

But you can probably just reduce the volume in your player by 3-6 dB and never have to worry about it.

This may indeed be the cause of the "harshness" that was being discussed here though, but I suspect it's a combination of things. Couldn't hurt to try though.

 

Allmost all 44.1/16bit files from Rock/Pop and similar Genres show values up to +1dB True Peak. Especially CDs from the 90th. Only very, very simplified and very low budget DACs will distort with True Peaks in this range. Some tracks may go up to +2dBTP or +3dBTP... and even higher (values above +1dBTP are typically NOT a result of heavy compression but of very fast changes in dynamics & attack in conjunction with max. digital values close to 0dBFS).

This usually happens when someone does peak level normalization on the peak level sample at some point in the mastering stage so that the album is as loud as it can possibly be.

 

There's no problem doing this in the digital domain, except that the actual output from DAC may be clipped, since the reconstructed waveform may extend higher than the 0 dBFS sample peak.

 

The really bad inter-sample peaks are found when converting from lossless to lossy formats. You need to reduce the volume when you encode to lossy formats, or else you risk creating significant inter-sample peaks. I have some tracks with peaks approaching +6dB!

 

A good DAC will handle this without issues...

I think it depends on the DAC. As far as I know, Benchmark are the only company which advertise that they can handle inter-sample peaks of up to +3.5dB with their DAC2.

Of course I am sure that most high quality DACs will be able to handle moderate inter-sample peaks without clipping, but as I said above it's easy enough to avoid them entirely by reducing the volume in your player.

[copy_of_a]

There's no problem doing this in the digital domain, except that the actual output from DAC may be clipped, since the reconstructed waveform may extend higher than the 0 dBFS sample peak.

... which is actually the short definition of Intersample Peaks :-)

 

The really bad inter-sample peaks are found when converting from lossless to lossy formats. You need to reduce the volume when you encode to lossy formats, or else you risk creating significant inter-sample peaks. I have some tracks with peaks approaching +6dB!

exactly. Also software based SRC in a Music Player may raise True Peaks significantly.

 

A good DAC will handle this without issues...

I think it depends on the DAC. As far as I know, Benchmark are the only company which advertise that they can handle inter-sample peaks of up to +3.5dB with their DAC2.

Actually I was referring to the True Peaks in the Ray Of Light track (short below +1dBTP). I think this souldn't be an issue for any up-to-date DAC to be used in a HiFi-System... regardless of the price. Otherwise it would be nearly impossible to listen to the vast majority of CDs (44.1kHz/16bit files respectively). The higher the True Peaks the more the quality of the DAC (its design respectively) may come into play... most likely that's true.

 

as I said above it's easy enough to avoid them entirely by reducing the volume in your player

... in case of doubt that's certainly sound advice.

[Matty70]

I purchased MusicScope last night and also got 10Euro coupon for any HIRESAUDIO download.

 

It is a very nice piece of software, but be aware that it does not analyze AIFF files.

 

Yep, discovered that after I purchased it. The developers say they are working on AIFF support though. it's fairly useless to me without AIFF support, but nice when you can use it.

 

Sorry for going back to off-topic for a short moment again. AIFF support ist now available in version 1.1.0 .

For further discussions about MusicScope I propose to use this thread.

[tgb]

hi,

to close this topic i opened : i found the reason why Madonna sounded harsh.

I use a Syno 212J with Minimserver, which streams to a sms100.

Yesterday i received the 12v13a LPSU from Teradak to power the nas... and Madonna sings so nicely now... great... analogic, fluid, dynamic, details. Big quality jump.

Sorry to you Madonna, you did a great job, the problem was on my side ;-)

[tranz]

Hi tgb,

 

yes I really like this tool very much. It gives you a very good insight into many quality aspects of music material.

What I like most is the flexibility of usage, the compatibility to many file types, including DSD and that it is implemented with the EBU R128 and ITU-R BS.1770-3 standards.

 

I'm not involved in any way in the company or in the devolopment of this software. It has been developed by a friend of mine.

So I think I'm one of the first users and I am really enthusiatic in using it, especially to check my highres downloads.

 

Regards

Matthias

 

Thanks for pointing out this software. I was looking for one with true peak and history, but the likes of Nugen were a bit too expensive for me. Any chance it will be released as a VST?

 

Cheers