Best chamber/small ensemble native hirez performances

[ted_b]

When it comes to classical music my tastes are much more like my jazz tastes: small acoustic combos. OK, not exactly classical-speak...more like chamber music and the small ensemble type recordings. String Quartets, cello concertos (in small quantities, just cuz I love the sound of cello), etc.

 

It seems we have some real good classical music listeners/downloaders here and I wondered where the smaller acoustically intimate gems are. I have some TACET DVD-Audio surround recordings of string quartet and quintets that are really nice, but their multichannel approach is quite aggressive, i.e one instrument per channel (they even have a 5.1 layer that rotates instruments around you during playback...!). Anyway...my point is that I love the smaller scale classical stuff and wondered what the sonic-and-performance gems are specially in download form (2 channel, of course). Thanks.

 

[Russell_L]

Hi Ted--A couple immediately come to mind:

 

Beethoven: Late Quartets - Cypress Quartet (96/24) (2 volumes):

 

http://cypressquartet.bandcamp.com/album/beethoven-late-quartets-vol-1-hi-res

http://cypressquartet.bandcamp.com/album/beethoven-late-quartets-vol-2-hi-res

 

Russian Cello Sonatas (Rachmaninoff/Borodin/Shostakovich) - Chaushian/Sudbin (88.2/24):

 

http://www.eclassical.com/labels/bis/russian-cello-sonatas.html

 

Russell

 

[Manelus]

Mozart - Divertimento Trio K563; Schubert String Trio - Trio Zimmermann (2011 BIS 24bit 88.2kHz) PCM eClassical

 

Mozart - Colloredo Serenade K.203, Divertimento K.251 - Janiczek {Linn 2008 24bit 88.2kHz} PCM Linn Records (not chamber, but small orchestra).

 

Schubert - String Q. 14 Death and the Maiden & 10 - Yggdrasil Quartet (BIS 2002 24bit 44.1kHz)PCM eClassical

 

String Quartets Beethoven, Nordheim, Bartok - Engegårdkvartetten (2010 2L 24bit 96KHz) 2L (DXD recordings, no conversion needed to PCM, clean spectrum and perfect sound)

 

Bach - Trio Sonatas - London Baroque (BIS 2002 24bit 44.1kHz) PCM eClassical

 

Mozart & Beethoven - Wind & Piano Quintets - Houg, Berlin Ph Wind Q (BIS 2007 24bit 44.1kHz) PCM eClassical

 

 

[ted_b]

seems to be easy on the wallet too. I'm really loving the Shostakovich (although not a symphony fan, his No. 5 is my all-time fave), but it's early.

 

Oh, and I have String Quartet 24/352.8k DXD sample from 2L's test bench site that I'll play directly with the Young DAC (384k capable). Look forward to it. Too bad they don't sell the whole album in native DXD.

 

[tubes59]

Eclassical's download manager is windows only at this time. I just learned this after paying for my download. Now I'll figure out what to do with what I just received without the manager...maybe it's no big deal.

 

[Russell_L]

It might be a slight pain without a download manager, but it's pretty straightforward to download each file individually from the list. (And their servers are very fast.) Just be sure to click on the "24 bit" link on each one (assuming you've bought the 24-bit files, that is). ;-)

 

Russell

 

[tubes59]

downloading each track one by one and getting the tracks into album organization did happen automatically (was worried about that). Having said that though, the download managers I've used (HDtracks and B&W) make the process MUCH easier, more autopilot like and faster. Hopefully eclassical will have a download manager for we Mac users soon.

 

[flatmap]

While bigger than a quartet, this chamber ensemble gives an excellent performance and wonderful recording: http://www.bso.org/bso/shop/audioDetail.jsp?pid=prod3880159&area=&id=bcat13360032

 

The ensemble is the Boston Symphony Chamber Players

composed of the just the first-chair string and wind players from the Boston Symphony Orchestra.

 

 

I chose the 24/88.2 AIFF for $9.99. They also offer WMA format (and SACD discs). Sample excerpts available on the download page.

 

[ted_b]

all-you-can-download subscription looks like a no-brainer? I wonder if they allow you to download BOTH the 2 channel and 5.1 channel versions of each (where available) if you have this seemingly unlimited (about 11 albums in hirez, half of them also in 5.1 hirez surround) 3 month pass. Thx

 

[flatmap]

I am so happy to see BSO do this; their albums frequently get awards, glowing reviews, and rightly so. With their digital store offering all new recordings in high resolution, they are way out in front of the pack.

 

I've gotten all of their high resolution offerings and a few of their mp3 archive recordings. Terrific.

 

[carelessethiopian]

The ENGARDKVARTETTEN Haydn string quartet Op. 76 no. 5 is incredible. (available i think both from HD tracks and Linn) When i want to show a Classical music fan what a high res digital file sounds like thats the one i play. I don't care for their Beethoven as much. I second the recommendation for the BSO Chamber Players Mozart.

 

I don't know if you consider it chamber music but the Dunnedin Consort Bach B minor mass (Available from Linn) is an amazing One voice per part performance on period instruments with a small (13 instrumentalists i think) ensemble. To me it sounds to be extremely well recorded.

 

[RobbieC]

Since 2L's test bench was mentioned, I was wondering if anyone has checked 2L's test bench DXD offerings for actual bit depth and sample frequency. I checked a few (actually, all of them) with Audacity and though I am by no means an authority I would have to conclude there is very little if any content above 22kHz on those DXD offerings.

 

I really like the concept of their test bench but I am not sure what to make of it if Audacity shows that spectral content for their DXD and lesser resolutions are the same.

 

[Mike in MD]

Ted,

 

Since you are also interested in jazz, how about starting another thread and listing some of your favorite jazz hiRez downloads? If you do, I will post my suggestion for Ron Carter and Rosa Passos: Entre Amigos (24/96) on HDtracks.

 

And to stay on topic, I strongly second the recommendation for

Dunedin Consort

J.S. Bach Mass in B Minor

at 24/88.2 at Linn

The spectrum is really clean and the music is great. A lot of it is small ensemble playing. If Bach had been alive in the 1950s, he would have been playing with Monk and Bird.

 

Mike

 

[The Computer Audiophile]

Testing attachment function for goldsdad :~)

 

[Mike in MD]

And look at the pretty noise bands going across the lower part of the spectrum.

 

Mike

 

[goldsdad]

Hi Rob

 

I downloaded three of the DXD resolution (24/352.8) WAVs - see Audacity spectrograms below. The images reveal only noise above 88.2kHz. Clearly, 24/176.4 recordings would have captured the audio frequencies.

 

Anyway, given that the recordings were made at 24/352.8, I think the main value of the 352.8kHz files is that they have not been resampled - they are the pure version of those particular recordings. The available 192kHz and 96kHz files derived from the DXD masters will be distorted unavoidably to some extent by the resampling process. I have no idea how noticeable these distortions would be when listening to the files.

Of course, very few people currently have playback systems that can avoid any resampling at all anywhere in the path from 352.8kHz digital file to analogue signal.

I guess the DXD files will still be interesting to people wanting to downsample the master to the optimum rate for their system, or comparing resampling methods such as Apple AUConverter and iZotope Resampler with different parameters.

 

 

 

 

 

 

 

[RobbieC]

My spectrogram for 2L38_01_DXD looks a little different from yours so perhaps my Audacity settings are not the preferred ones. I've attached a screen shot of my result below.

 

Can someone correct me if I do not have this right:

 

In order to create a 176.4 kHz file from the originally captured 352.8 kHz

file, the DXD file must first be low-pass filtered (anti-alias) and then have

every other sample discarded.

 

Is it distortion resulting from the required low-pass filtering that you mean? Is there audible distortion with this kind of low-pass filtering? Do you mean distortion related to Gibbs phenomenon due to loss of high frequency content or distortion from phase shifts related to the slope of the roll off of the filter?

 

(Sorry, I just graduated last year with a degree in mathematics but specialized in differential geometry so am not fully versed in digital filters or Fourier. I'll try not to let my contribution to this board stray to far toward the rigorous.)

 

Also, why is there a relatively "clean" gap of no content between what is presumably the performance content area and that area in the upper frequencies we are calling "noise"? Is it a limitation of the microphones used or perhaps a limitation somewhere else in the recording chain?

 

I recall Cookie, I believe, saying that sometimes the banding and "noise" in a spectrogram is from air conditioners in the background, reed slap, spit, etc. Speaking of reeds, reeds from my understanding approach square wave motion by design so one would expect a lot of high frequency content coming from these instruments. And for that matter, string instruments operate by rosined bows repetitively grabbing and suddenly releasing strings, a process that approaches sawtooth wave motion. They should have just as much if not more high frequency content than reed instruments. I say that because sawtooth waveforms are essentially some fundamental plus all its harmonics, respective amplitudes multiplied by the inverse of harmonic number of course, while square waveforms include only the odd harmonics. These harmonics should extend I imagine several octaves up such that if a violin could produce C8, as on a piano with a fundamental at 4186 Hz, just going up five octaves gets you to 134 kHz. Where is that content? Why is it not showing up? Again, microphones?

 

[Mike in MD]

Rob,

 

Goldsdad has the gain cranked up so the lower frequencies are red on a blue background. I like your settings better. You can play with the preferences window. If you raise "Frequency gain (db/dec)" you will see more high freq extension. I have it set to 5. Your spectrogram looks really nice. I can't answer your technical questions about the effects of filtering. When you figure it out, you can start a new thread on that.

But... now I know someone to contact when I run into trouble on my readings in Lie Groups. :}

 

Mike

 

 

[Mike in MD]

I recently downloaded

Ensemble Meridiana

Tastes of Europe: Telemann Trios & Quartets

at Linn

 

I got the 24/88.2

They also offer 24/192. That seems wrong. Either it should be 96 and 192, or 88.2 and 176.4 Jared said that Linn masters in PCM, but this seems to be some kind of conversion. But whatever.

 

The spectrum is very clean with lots of high frequencies (characteristic of Linn) and I like the music. Not everyone likes baroque chamber music.

 

Mike

 

 

[goldsdad]

"My spectrogram for 2L38_01_DXD looks a little different from yours so perhaps my Audacity settings are not the preferred ones"

 

The preferences for my spectrograms was as in the image below.

 

 

 

Window size is 4096 because the larger the value, the more precise the frequency computation (at the expense of time precision).

 

As you have discovered, the box for Display Maximum Frequency has a limit of 100kHz. I normally leave it at 96kHz for files of all sample rates because the max. that actually will be displayed never exceeds a file's Nyquist freq.

The value entered is the max. freq. that will be displayed when a file's spectrogram is calculated initially. To see a higher range when a file's Nyquist freq. exceeds that limit, right-click (i.e. vertical zoom out) in the column of freq. labels to the left of a graph.

 

The Display Range is 144dB because that is the dynamic range of a 24-bit file (i.e. 6dB/bit) and I was interested in the entire range including the lowest-level noise pattern, and seeing the frequency extension of the music at all levels, right down to the noise floor.

 

 

"Can someone correct me if I do not have this right:

In order to create a 176.4 kHz file from the originally captured 352.8 kHz

file, the DXD file must first be low-pass filtered (anti-alias) and then have

every other sample discarded."

 

I think an expert on resampling, which certainly rules out me, should step in here to explain some of the various resampling methods.

 

 

" Is it distortion resulting from the required low-pass filtering that you mean? Is there audible distortion with this kind of low-pass filtering? Do you mean distortion related to Gibbs phenomenon due to loss of high frequency content or distortion from phase shifts related to the slope of the roll off of the filter?

(Sorry, I just graduated last year with a degree in mathematics but specialized in differential geometry so am not fully versed in digital filters or Fourier. I'll try not to let my contribution to this board stray to far toward the rigorous.)"

 

Distortions of which I'm aware include pre-ringing, post-ringing and phase shifts. Again, we could use an expert here. As I said earlier, I don't know how audible these distortions are likely to be for any particular listener.

 

 

"Also, why is there a relatively "clean" gap of no content between what is presumably the performance content area and that area in the upper frequencies we are calling "noise"? Is it a limitation of the microphones used or perhaps a limitation somewhere else in the recording chain?

I recall Cookie, I believe, saying that sometimes the banding and "noise" in a spectrogram is from air conditioners in the background, reed slap, spit, etc. Speaking of reeds, reeds from my understanding approach square wave motion by design so one would expect a lot of high frequency content coming from these instruments. And for that matter, string instruments operate by rosined bows repetitively grabbing and suddenly releasing strings, a process that approaches sawtooth wave motion. They should have just as much if not more high frequency content than reed instruments. I say that because sawtooth waveforms are essentially some fundamental plus all its harmonics, respective amplitudes multiplied by the inverse of harmonic number of course, while square waveforms include only the odd harmonics. These harmonics should extend I imagine several octaves up such that if a violin could produce C8, as on a piano with a fundamental at 4186 Hz, just going up five octaves gets you to 134 kHz. Where is that content? Why is it not showing up? Again, microphones?"

 

Microphones do, of course, have frequency limits. I doubt any used in recording musical performances will exceed 100kHz, and most models will stop significantly lower. Also, the decreasing amplitudes of higher harmonics at freqs. within the mic. limits will result in their being lost in either the ambient noise of the performance or the noise floor of the recording chain.

 

As seen in my images, but lost in yours, due to our different range setting in the preferences, noise permeates the entire recording. On closer analysis of track 2L38 just now, even in the region of the noise's lowest amplitude, it appears to consume perhaps the 6 least significant bits of each sample.

 

 

[goldsdad]

Hi Mike

 

"Goldsdad has the gain cranked up so the lower frequencies are red on a blue background"

 

No, I didn't have the gain cranked up at all - it was at 0dB.

 

I did have the display range at 144dB, not because it will result in a red on blue display, but in order to view information in the full dynamic range of a 24-bit file, regardless of audibility.

 

 

[Bob Stern]

@Ted:

 

The reason that I, and perhaps most, classical aficionados have a hard time recommending chamber music in hi-res is that almost none of the great performances I'm familiar with have been released in hi-res.

 

However, if you are relatively new to classical chamber music, so that you aren't already familiar with the "basic repertoire" or "masterworks" of classical chamber music, perhaps I and others can recommend those works to you. You then can search for what hi-res recordings of those works are available.

 

Schubert's "Death and the Maiden" string quartet, recommended above, certainly is one of the masterworks, although I haven't heard of those performers.

 

I strongly endorse Russell's recommendation of the Cypress String Quartet. However, if you haven't been exposed to Beethoven's late quartets, Beethoven's op. 59 quartets are much easier to grasp.

 

Wispelwey & Lazic have a great Channel Classics recording of cello sonatas by Shostakovich, Prokofiev and Britten, although the Britten is not my cup of tea.

 

I dislike Wispelwey's Brahms sonata recording. He completely lacks the romantic temperament for Brahms.

 

[RobbieC]

Thanks Gold's dad. It was indeed that my range was set at 96 dB.

 

Not sure what this setting is doing with colors though.

 

 

Copied from the Audacity manual (Spectrogram Preferences):

 

Range (dB): Affects the range of signal sizes that will be displayed as colours. The default is 80dB and means that you won't see anything for signals 80dB below the value set for 'Range'. This option has no effect on the Pitch (EAC) view.

 

I'm not sure what this is saying. The default value set for 'Range' is 80 dB, right? So if I left it at the default, I won't see anything for signals 80 dB (this default value) below 80 dB (again, that same default value set in the 'Range' box)? What?

 

[goldsdad]

Well spotted. It might be better to say:

 

Range (dB): Affects the range of signal sizes that will be displayed as colours. The default is 80dB, and means that no colour will be displayed for signals below -80dBFS. This option has no effect on the Pitch (EAC) view.

 

 

[TimArruda]

Russell,

 

Thanks for the recommendation of Cypress Hill Quartet. I downloaded Vol. 1 of the Beethoven Late Quartets in 24/96 FLAC. I'm listening to them right now via my MacBook Pro -> Decibel -> USB -> Burson Audio HA-160D -> Sennheiser HD500 (waiting on a pair of Audez'e LCD-2's) and they sound fabulous.

 

Tim