Geek
Headphoneus Supremus
- Joined
- May 5, 2002
- Posts
- 1,810
- Likes
- 10
My system:
SOURCE - RME 96/8 PAD, analog out
INTERCONNECT - Dimarzio M-path
AMP - Headroom Maxed Out Home Reference + stepped attenuator
HEADPHONE - HD650 + cardas 10' cable
I finally received the Stereophile Editor's Choice CD in the mail after 3 months of waiting. I have heard many of the test tracks on incomprehensibly expensive systems. When I ripped the songs to FLAC using the highest possible quality standards in EAC, here's what I found out:
Using the RME at 44.1khz, the sound was rather disappointing from an audiophile perspective. For one, bass was a little sloppy and not as deep as it could have been on the fender bass guitar tests. Otherwise the sound was acceptable but not compared to a good CD player.
The midrange was probably where the majority of my troubles occured. Midrange in my opinion is something absolutely essential for the music to shine, and if it's not done right then the sound is ruined to my ears. Midrange is also extraordinarily difficult to get right in the audio world; even some really good high end phones don't have a strong midrange.
When I listened to the Mozart clarinet quintet, my eyes literally shot wide open at the sonic misfortune of the RME's midrange.
Let me explain a little background here for those of you who do not know me. I am an avid, obsessed violinist and musician and am currently enrolled in a college music program. During high school, our chamber orchestra (we have a near-professional quality chamber group as well as an equally excellent symphony at our high school) played this same concerto that's on the CD with an incredible clarinetist, Dr. Gregory Young. We rehearsed the piece for a total of about 20 hours. It was really truly amazing.
Now, I return to the recording and how it sounded on my setup. The soundstage is literally 35 degrees wide and it exhibits something us musicians like to call the "cavern effect," in which it sounds like the performance is there but we're listening it through a narrow tunnel due to echoy resonances and colorations that shouldn't be there. All of these colorations were in the midrange. The cello lacked body and proper texture, instead tending to ring loudly on open C and D, glaring errors which I'm sure stereophile would never allow on one of their recordings. Violin has no midrange to speak of on my setup, which is saddening; I like the violin's midrange and wanted to hear it. Violas disappear, tonally speaking. I love midrange detail, and I want to hear the tonal center of each instrument which is lower midrange for classical music (Cello A = 220hz, Viola/violin A=440hz). There's plenty of highs, granted, and really powerful lows, but I'm interested in the body of the music.
I cannot listen to this track at volumes approaching a live performance without literally RINGING in the midrange overwhelming the music! Not a good thing for the RME card.
Now on to highs. The highs of stringed instruments and cymbals/hats are something I've never heard a digital system reproduce. NEVER. EVER. I haven't heard the Meridian GO7/GO8 series yet, and Todd claims they're really close to vinyl in smoothness (I define smoothness as not distorting the highs and having a tonally accurate sound). But the RME 96/8 does not break any barriers here. Actually, it represents a giant leap backwards.
Whilst not sounding worse than the hashy grainy crap that is my santa cruz card trying to play a cymbal, the sound isn't much better. Going back to the Clarinet Quintet, we have some bad thigns happening. When a string player draws his bow across the string, a fundamental vibration (say, 440hz) is produced. On top of that, two halves of the string vibrate as well, producing what is called a harmonic, which is on top of the root 440hz, or 880hz. Then these two halves can be divided further into 1.76khz quarters, and so on. The first harmonic is going to sound a lot less loud than the root, the second harmonic less loud than the first, and so on. Bow noise is another factor. When a string vibrates, it does not do so in such a way that produces a sine wave. Rather, it is gripped by the rough texture of the bow, released, and gripped again, hundreds of times per second during a normal smooth bow stroke. This creates that sound unique to a bowed instrument. Now add bow noise, giving the instrument that slightly brittle sound.
When you throw all of these together, add in the sonic properties of the instrument's sounding box and room, you get what I like to call an audiophile's worst nightmare. It's devilishly difficult to accurately reproduce this, which I should term "string texture," and something that primarily fails to be done in the highs. This texture is a combination of harmonics, bow grip, and bow noise. If you're intimately familiar with how it should sound like, you'll never heard it done perfectly on a recording, I guarantee it.
I fully expect this when I am about to listen to a new recording. However, I was startled by the complete lack of any detail whatsoever above what is probably the 10 or 12khz point. I heard a little bit of proper texture, something that keeps my brain thinking, "yes, these are stringed instruments." But the other ninety percent of my mind was thinking "what the heck is that fuzz!?". I have to explain what I heard: most of the highs are literally "fuzzy," lacking any detail whatsoever. Where you hear lots of fuzz, you know that originally there was something really neat like a loud, aggressive note or mabye that was the performer breathing. But you really don't ever get to know, because it's just fuzz. When I should be hearing the clarinetist tapping his pads, I hear a bit of that in the mids (the more visceral "pop,") but there is a lack of highs, rather, it's just fuzz. Lots of it. And it gets really annoying, so bad that I just switched off foobar after a few minutes of listening and went to the next track.
On to how accurate the highs are on cymbals. On track 15, Ellington, arr. Harris: "The Mooche," we have a nice jazz arrangement; featuring a nice setup of instruments and above all some really clear drums. This track I have had the fortune of hearing on all kinds of setups at headroom.
I think that it's a safe bet to say that an iriver MP3 CD player can do the cymbals better than the RME 96/8 with my maxed out home. Doubting that the maxed out home is at fault (it knocked my socks completely off hooked up to a bluenote vinyl rig), I place the majority of the blame on the 96/8. Cymbals are distorted. If my ears had a butt, that butt was seriously kicked, and kicked hard after hearing that song.
The RME does have some redeeming qualities, however, in spite of this massive angry negatively biased rant that I wrote. It's really really versatile. It supports 96khz oversampling, but I hear no difference between 44.1 and 96 sample rates (44.1 sounds slightly cleaner, if anything), since all of the music I own is pretty much recorded to 44.1/16 and it doesn't seem to make a difference resampling. (note that the resampler in foobar comes with a note that sound quality will not be improved upon by higher sample rates. I have come to agree with this, there isn't a difference at all). If you are recording lots of tracks, the card is going to be great, but if you feed the recording through that card on the way out, congrats, you've just created garbage. Otherwise, the card does a great job when it comes to using professional mixing and music creation programs such as cooledit and fruity loops producer edition. I honestly question its $300 price tag, however. My money would've been more well-spent on flashy colored lights on my case, or some sets of strings for my violin, or perhaps an iriver PDCP, something I need badly now but have no money to afford.
If you like classical music, and you like it a lot, do yourself a favor and avoid this sound card. Also avoid the M-audio revolution, the turtle beach santa cruz, and all of the "gaming" cards out there. Basically dodge PC audio unless you don't care about sound quality and archive all of your music digitally in an easy to reach fashion. What you really need to do is get a standalone CD player. There are a ton of manufacturers out there, and all of them seem to offer a better bang for the buck when it comes to sound quality.
Cheers,
Geek
SOURCE - RME 96/8 PAD, analog out
INTERCONNECT - Dimarzio M-path
AMP - Headroom Maxed Out Home Reference + stepped attenuator
HEADPHONE - HD650 + cardas 10' cable
I finally received the Stereophile Editor's Choice CD in the mail after 3 months of waiting. I have heard many of the test tracks on incomprehensibly expensive systems. When I ripped the songs to FLAC using the highest possible quality standards in EAC, here's what I found out:
Using the RME at 44.1khz, the sound was rather disappointing from an audiophile perspective. For one, bass was a little sloppy and not as deep as it could have been on the fender bass guitar tests. Otherwise the sound was acceptable but not compared to a good CD player.
The midrange was probably where the majority of my troubles occured. Midrange in my opinion is something absolutely essential for the music to shine, and if it's not done right then the sound is ruined to my ears. Midrange is also extraordinarily difficult to get right in the audio world; even some really good high end phones don't have a strong midrange.
When I listened to the Mozart clarinet quintet, my eyes literally shot wide open at the sonic misfortune of the RME's midrange.
Let me explain a little background here for those of you who do not know me. I am an avid, obsessed violinist and musician and am currently enrolled in a college music program. During high school, our chamber orchestra (we have a near-professional quality chamber group as well as an equally excellent symphony at our high school) played this same concerto that's on the CD with an incredible clarinetist, Dr. Gregory Young. We rehearsed the piece for a total of about 20 hours. It was really truly amazing.
Now, I return to the recording and how it sounded on my setup. The soundstage is literally 35 degrees wide and it exhibits something us musicians like to call the "cavern effect," in which it sounds like the performance is there but we're listening it through a narrow tunnel due to echoy resonances and colorations that shouldn't be there. All of these colorations were in the midrange. The cello lacked body and proper texture, instead tending to ring loudly on open C and D, glaring errors which I'm sure stereophile would never allow on one of their recordings. Violin has no midrange to speak of on my setup, which is saddening; I like the violin's midrange and wanted to hear it. Violas disappear, tonally speaking. I love midrange detail, and I want to hear the tonal center of each instrument which is lower midrange for classical music (Cello A = 220hz, Viola/violin A=440hz). There's plenty of highs, granted, and really powerful lows, but I'm interested in the body of the music.
I cannot listen to this track at volumes approaching a live performance without literally RINGING in the midrange overwhelming the music! Not a good thing for the RME card.
Now on to highs. The highs of stringed instruments and cymbals/hats are something I've never heard a digital system reproduce. NEVER. EVER. I haven't heard the Meridian GO7/GO8 series yet, and Todd claims they're really close to vinyl in smoothness (I define smoothness as not distorting the highs and having a tonally accurate sound). But the RME 96/8 does not break any barriers here. Actually, it represents a giant leap backwards.
Whilst not sounding worse than the hashy grainy crap that is my santa cruz card trying to play a cymbal, the sound isn't much better. Going back to the Clarinet Quintet, we have some bad thigns happening. When a string player draws his bow across the string, a fundamental vibration (say, 440hz) is produced. On top of that, two halves of the string vibrate as well, producing what is called a harmonic, which is on top of the root 440hz, or 880hz. Then these two halves can be divided further into 1.76khz quarters, and so on. The first harmonic is going to sound a lot less loud than the root, the second harmonic less loud than the first, and so on. Bow noise is another factor. When a string vibrates, it does not do so in such a way that produces a sine wave. Rather, it is gripped by the rough texture of the bow, released, and gripped again, hundreds of times per second during a normal smooth bow stroke. This creates that sound unique to a bowed instrument. Now add bow noise, giving the instrument that slightly brittle sound.
When you throw all of these together, add in the sonic properties of the instrument's sounding box and room, you get what I like to call an audiophile's worst nightmare. It's devilishly difficult to accurately reproduce this, which I should term "string texture," and something that primarily fails to be done in the highs. This texture is a combination of harmonics, bow grip, and bow noise. If you're intimately familiar with how it should sound like, you'll never heard it done perfectly on a recording, I guarantee it.
I fully expect this when I am about to listen to a new recording. However, I was startled by the complete lack of any detail whatsoever above what is probably the 10 or 12khz point. I heard a little bit of proper texture, something that keeps my brain thinking, "yes, these are stringed instruments." But the other ninety percent of my mind was thinking "what the heck is that fuzz!?". I have to explain what I heard: most of the highs are literally "fuzzy," lacking any detail whatsoever. Where you hear lots of fuzz, you know that originally there was something really neat like a loud, aggressive note or mabye that was the performer breathing. But you really don't ever get to know, because it's just fuzz. When I should be hearing the clarinetist tapping his pads, I hear a bit of that in the mids (the more visceral "pop,") but there is a lack of highs, rather, it's just fuzz. Lots of it. And it gets really annoying, so bad that I just switched off foobar after a few minutes of listening and went to the next track.
On to how accurate the highs are on cymbals. On track 15, Ellington, arr. Harris: "The Mooche," we have a nice jazz arrangement; featuring a nice setup of instruments and above all some really clear drums. This track I have had the fortune of hearing on all kinds of setups at headroom.
I think that it's a safe bet to say that an iriver MP3 CD player can do the cymbals better than the RME 96/8 with my maxed out home. Doubting that the maxed out home is at fault (it knocked my socks completely off hooked up to a bluenote vinyl rig), I place the majority of the blame on the 96/8. Cymbals are distorted. If my ears had a butt, that butt was seriously kicked, and kicked hard after hearing that song.
The RME does have some redeeming qualities, however, in spite of this massive angry negatively biased rant that I wrote. It's really really versatile. It supports 96khz oversampling, but I hear no difference between 44.1 and 96 sample rates (44.1 sounds slightly cleaner, if anything), since all of the music I own is pretty much recorded to 44.1/16 and it doesn't seem to make a difference resampling. (note that the resampler in foobar comes with a note that sound quality will not be improved upon by higher sample rates. I have come to agree with this, there isn't a difference at all). If you are recording lots of tracks, the card is going to be great, but if you feed the recording through that card on the way out, congrats, you've just created garbage. Otherwise, the card does a great job when it comes to using professional mixing and music creation programs such as cooledit and fruity loops producer edition. I honestly question its $300 price tag, however. My money would've been more well-spent on flashy colored lights on my case, or some sets of strings for my violin, or perhaps an iriver PDCP, something I need badly now but have no money to afford.
If you like classical music, and you like it a lot, do yourself a favor and avoid this sound card. Also avoid the M-audio revolution, the turtle beach santa cruz, and all of the "gaming" cards out there. Basically dodge PC audio unless you don't care about sound quality and archive all of your music digitally in an easy to reach fashion. What you really need to do is get a standalone CD player. There are a ton of manufacturers out there, and all of them seem to offer a better bang for the buck when it comes to sound quality.
Cheers,
Geek