What I want in a DAC

[jrebman]

Hi John,

Thanks for the reply and that sounds perfect to me.

I don't know how things have progressed since last RMAF, but I was really impressed with what I heard of the dac at that point -- even more so when I later learned that it was being fed directly from the optical out on the Mac Mini.

Now I'm really looking forward to all of this!

-- Jim

 

[Yoder]

... my guess would be that the only source suitable for conversion to 24/192 resolution would be the masters or copies of them. I have found that alot of the files available at HD Tracks and others are vinyl rips.

As I understand it they use 24/192 for the master in digital recording now days, though some are probably beginning to push that envelop, and are starting to push 24/96 as the consumer standard. Though Apple's iTunes Plus will present a different story about hi-res mp3's.

You sure about the vinyl rips? Most re-masteriing uses the original master tape (There are some good movies on Netflix under "music" called "Classic Albums." You will see Steely Dan, the Dead, etc. using what looks like 24 track tapes and some 8 track from way back.)

I have been doing some vinyl ripping of my own, using a borrowed Benchmark ADC. I was told that converting vinyl to anything above redbook was a waste of time, but I found ripping to 24/96 gives me the best results.

Somebody does not know what they are talking about. Going above Redbook would only be a waste if your intent was to burn all of your vinyl music to CDs. If you have the hardware and software to handle 24/96, or better yet 24/192, and plan on using computer playback then go for it. 24-bit is far superior to 16-bit.

I noticed NAD has released a DAC that is capable of 32 bit / 384 khz resolution. What for? Are they hoping that at some point the recording studios will begin producing hi-rez along with other formats for general consumption? Just wondering if these capabilities available now are worth spending money for.

It is just a numbers game. I have not heard any advantage in up sampling beyond the native rate. That being said, I have not even seen any 32 bit audio for sale. Regarding the purchasing of such, it is NEVER cheaper to buy new technologies. Personally, I don't thing going beyond 24/96 would make much difference to my aged/tired ear drums, and so I am perfectly content with 24/94.

 

[earwaxxer]

Interesting discussion.... What I might add to the mix is that, IMO and experience, there are potentially other factors that influence the sound at various sampling rates/resolutions. What I have found is that my laptop upsampled redbook (24/96) using Sox and minimum phase settings (0%) sounds the best, better than the usual default linear phase (50%). This is a curious result because my Transporter upsamples redbook in realtime with Sox as well, but passes high res. without upsampling. I have tried various phase settings from linear to minimum. All of them sound better than native redbook, minimum phase is the most 'vinyl like' IMO. This may have to do with filter effects on ring/pre-ring.

So, in this case, redbook that is upsampled before hand appears to sounds superior to redbook that is passed to the DAC as 16/44.1 and upsampled by the DAC. This is not a small effect. It is very clearly different. Everytime I get to doubting myself and switch back to native redbooks rips I come to the same conclusion. I wish it was not true, because I could have my entire 1000 CD collection on my laptop in native rips. Now I have to settle for 250 (large upsampled) albums or so. I would quantify the differences I hear in the area of 'ease and liquidity'. A 'naturalness'. Musical and believable with an added lack of fatigue and irritation. Very clear and obvious.

This effect has also narrowed the gap as far as my system is concerned, if there even is one, between native HD tracks 24/96 downloads and my upsampled redbook. I have not found that high res. downloads sound any better than a good redbook recordings massaged with Sox.

Now the question that is raised in my mind, is what role the various 'decoding' (filtering) processes have on sound quality, irrespective of resolution. Sorry if I'm making things more complicated! The plot thickens!

 

[earwaxxer]

Interesting new article on the subject...

http://www.enjoythemusic.com/magazine/manufacture/0512/

 

[Yoder]

Interesting discussion....

Interesting indeed...

I run my audio through Amarra that uses only the native rate, and fixed point. The difference between Amarra and iTunes is profound. (They have a little button one can use to make comparisons.) Use the Amarra parametric equalizer and the difference is even more dramatic. I feed my streaming digital audio from my mini into a 24-bit tube DAC and so really can't provide any feedback as far as 16 vs 24 bit in the DAC goes, but when I play 24/48 or 24/44 files through the system the difference it is like night or day when compared to Redbook specs. The key, for me, seems to be in the bit-depth. I have an external up sampler/clock that allows me to flip from 44.1 to 96 on the fly, and I can sit in front of the speakers and flip 24/44 files between 44.1 and 96 all day long and do not notice any difference. I am going to order a couple of SACDs today (24/96 and 24/192) and will try the same thing. My guess is that there will be a noticeable difference when I downsample a 96 or 192 file to 44. When upsampling the sample rate all that is happening is that a bunch of zeros are being tacked on, and my guess is that that is why I do not hear any difference.

There are some transports/DACs that advertise as having various capabilites using high-end DSPs, and adaptive time filtering "that applies polynomial curve-fitting interpolation that allows for greater data buffering and re-clocking" so as to remove "almost" all digital jitter. For me this creates a conundrum. Namely, all things being equal should I per sue higher bit-rates that that will sound great when played through excellent gear at the native rate, or is it worth per suing high-end DSPs/hardware and the improved algorithms that come with them (firmware that is actually altering the input audio in various ways)?

Personally, I am very content with going for hi-res audio and playing it at its native sample rate. Correct me if I am wrong, but this is pure, unaltered audio versus audio that has had various algorithms applied to it. Granted, algorithms are used in all digital playback but there seems to be a point where when applying these algorithms we go from decoding the digital signal to actually altering it in various ways from its original form in order to create a "better" sound.

Anyway, 24/48 audio blows me away as it is and I can hardly wait to hear the 24/96 I have on order. I will let you know if I hear any difference when I down sample the sample rate. The two SACDs I am getting are: Pink Floyd-Dark Side of the Moon, and The Band-Cahoots.

 

[Noskipallwd]

Hello, Yoder and Eric, thanks for the discussion. First off, I should edit the statement about HD Tracks. There seems to be more natively recorded hi-res files available. On many of the early downloads from HD Tracks, and other sites you can hear the vinyl rumble. Very faint, but there nonetheless. It seems there are more recording companies willing to supply Hi-res digital. I made some downloads from Harmonia Mundi and a Norweigan company called 2L,(who also have files available on their website), that sound absolutely fantastic. I am using Pure Vinyl to do my rips from vinyl and I am very happy with the results. I still prefer a pure analog vinyl rig but I think we are making progress.

Cheers,
Shawn

 

[earwaxxer]

Hey Yoder - Thanks for verifying that I am not the only crazy MF out there obsessing over this stuff!

Its pretty cool that your system gives you that much A/B control to be able to make some comparisons on the fly. I dont have that. I am convinced that shifting the aliasing etc. to higher frequencies with higher sample rates is key. Probably equally as important the filtering algorithms that come into play at the different sample rates. From my systems perspective, I'm not hearing a huge difference with native high res. material. It has more to do with the recording quality. A good example are the 2009 Beatles redbook remasters. Upconverted with Sox they sound as good or better than the native Paul McCartney 'Band on the Run' 24/96 HD Tracks download.

Cheers - Eric

 

[Yoder]

On many of the early downloads from HD Tracks, and other sites you can hear the vinyl rumble. Very faint, but there nonetheless.

If there is true remastering of older audio going on then they are using reel-to-reel tapes. Only with the original master tape can the sound engineer go in and play with the various tracks on the tape. The movies I referred to above go into a lot of detail about this and they play with eliminating various instruments, enhancing them, etc. This type of work can only be done with the original master tape and not vinyl or with any of the digital codecs (AiFF, WAV,mp3, mv4, etc.) once the tracks have been "compressed" from 24-tracks to stereo. The "rumble" you hear is probably the result of a poor quality master tape. Finding mint perfect master tapes seems to the real challenge today.

Here are a couple of quotes about some HDTracks hi-res audio. "...with our exclusive Rolling Stones release, ABKCO Records undertook a long and painstaking re-mastering process employing state-of-the-art master tape to digital file transfers using the best in Analog to Digital converters to provide HDtracks with the highest quality music files possible."

The engineer for Cat Stevens said:

 

[Doc B]

Not all masters are multitrack and thus you can't make that generalization. The term for mixing multitrack session tapes to stereo is mixing, not compressing. A bad master tape will not have lathe rumble. That is an artifact of cutting lacquer or playback on a noisy turntable. A bad master tape can have a lot of other problems - dropouts, rolled off highs, dulled transients, stretched spots and splices, sticky shed, maybe even "rocks", but not lathe or turntable rumble.

 

[Noskipallwd]

I have a download of Nojima plays Listz that you can hear the stylus hit the record, you can only hear it with volume up above normal listening levels. I have listened to enough vinyl in the last 30 some odd years to recognize that sound, and I am not the only person to notice this on various tracks. Please, don't get me wrong there is nothing wrong with the files quality, the Nojima piece sounds great. I simply, in my own lack of experience,  wonder how this business works. Where exactly is the media originating? Are we paying for files prepared by professionals with studio equipment and backing, or something done half-assed in somebody's basement? (similar to mine)!

Cheers,
Shawn

 

[Doc B]

I doubt very much that Nojima plays Liszt would have been transferred from a vinyl pressing to a high res file for commercial sale. Our Tape Project album of that title has been made from the original tape recording - which is in very good condition. AFAIK it was recorded simultaneously in 176/24, so there would be little reason to waste time transferring it from an LP.

Your description of the level being set above normal listening levels to hear the effect makes me think you are hearing the tape hiss starting as the tape was rolled (if it was transferred from the analog master) or possibly air conditioning rumble in the noise floor.

 

[earwaxxer]

Good point Shawn!

That is exactly my sentiment when I think about purchasing a 'new'  'high res' download. If there was all this effort placed in remix/remaster etc, etc, I would think that the studio would also be interested in releasing it in a blu-ray, DVD, CD, box set  etc. for mass distribution. Anyone can, and do (illegally), upload a vinyl high res. 'remaster'. When buying a new remaster I take into account who did the work (ex. Rhino) and what reviewers are saying about it.

 

[Noskipallwd]

Your description of the level being set above normal listening levels to hear the effect makes me think you are hearing the tape hiss starting as the tape was rolled (if it was transferred from the analog master) or possibly air conditioning rumble in the noise floor.

So many avenues for noise to get into the process, no wonder sound engineers tend to have obsessive personalities. I would love to hear the Tape Project version of Nojima Plays Liszt, it is a favorite of mine. I think you are right about the noise, the sound quality is darn good at 24/96, not as good as my vinyl version but good enough to listen to regularly. Maybe I should quit thinking about it and just listen!

Cheers,
Shawn

 

[Yoder]

Dang, three high-and-inside fast balls!

Not all masters are multitrack and thus you can't make that generalization.

True, but when talking about high-resolution audio don't you think it is a fair assumption that they are using multi-track studio tape masters?

The term for mixing multitrack session tapes to stereo is mixing, not compressing.

After years of working with Logic, I was taught that the term used to combine several tracks into mono or stereo is "bounce;" it is also a standard term in audio. Likewise, "mixing" was a term used to described the process of shaping the overall sound of the project by adjusting vol levels, panning, messing with the EQ, and using other effects on the project. But, being half brain dead and suffering from tip-of-the-tongue phenomenon at the time of the writing, all I could think of was compression, which actually deals with the "attack" and "release" in audio. So, I used the computer file definition of compression and related it to making a file smaller as in going from a huge multi-track project to a smaller file like an AIFF.

A bad master tape will not have lathe rumble.

Oops, forgot a word--perceived. 'The perceived "rumble" you hear is probably the result of a poor quality master tape.' In other words, I was questioning if the perceived "rumble"  was some other sound on the tape and was being mistaken as "rumble." In any case, I was suggesting that the poor sound was due to the tape.

Personally, I do not see how a company like HDTracks could burn "hi-res" audio off of vinyl and re-sale it as being "remastered." There are some crappy re-masterings out there, but they usually seem to be casualties of the "loudness wars," or as if someone went out for coffee and put the process on auto-pilot and allowed over clipping to take place.

@earwaxer: I got my hybrid two hybrid SACDs the other day. Looks like SACD uses the same licensing scheme as HDMI does. Ergo, SACD audio can only be played through analog outs. As soon as I change the output to my SPDIF or optical the player reverts to CD mode. Then it is in big print in the manual that "SACD can only be used with analog outs." Bad news is that one is reduced to using the DAC in the SACD player, but if one wanted a better DAC then there are always the high-end CD/SACD players available. So, I was unable to see if there is a much difference in down-sampling from 96 to 44. The good news is that I can still burn the Redbook CD to my hard drive.

 

[Doc B]

I think that companies that are reissuing stereo releases in high res are most often working from stereo mixes rather than multitracks. Bouncing is that act of mixing two or more tracks together onto an unused track so that the original tracks can be repurposed.

 

[jrebman]

Yes, what Doc says -- most of these companies are using stereo masters even though sometimes they use differently mixed masters from the album that is in general circulation  -- there is a thread on the Hi-rez music circle on audio circle that has a list of bad or alternate mixes being used by HD Tracks and others and people being mad about the fact that they don't sound like the same album they were expecting, but in hi-res.

As for SACD,it's not so much the licensing scheme as it is the underlying technology.  SACD is DSD encoding (Direct Streaming Digital), not PCM so there is only a streaming rate spec, not a bit-depth (as it is all one-bit depth)  This is in contrast to what Ryan Mintz of Core Audio seeems to be implying in his article in ETM that was mentioned earlier, and in fact, one-bit digital is here and has been here for a while and is obviously within reach of many.  Furthermore you can almost always bet that if you see a file on HDTracks that only comes in 24/88.2  PCM format, it was made from an SACD/DSD master as that is the PCM limit for commercial SACD recordings.

As a side note, Pure Music can natively play DSD files if the dac it is feeding has a DSD chip, but it can also take dDSD files and convert them on-the-fly to PCM format and play them on your normal PCM dac.

From my subjective perspective, I have a 2-layer SACD of Jorma Kaukonen and friends' "Blue Country Heart" and I'd far rather listen to the redbook layer through a good PCM dac than the sacd layer -- just sounds more like real music to me.

Several people have conducted a/b/c tests with vinyl, redbook and SACD, and most people preferred them in that order with SACD a fairly distant third.

BTW, there are ways to rip the DSD layer of an SACD to  PCM, but it does take some special software and a special disc drive, but not having done it myself, and only owning a very small handful of SACDs, I don't plan to to this myself -- especially when PM can do it for me.

-- Jim

 

[Yoder]

From my subjective perspective, I have a 2-layer SACD of Jorma Kaukonen and friends' "Blue Country Heart" and I'd far rather listen to the redbook layer through a good PCM dac than the sacd layer -- just sounds more like real music to me.

That is a good CD, though I think "Quah" is his all time classic and is on my top ten list.

I burned the CD version of "Cahoots" onto my hd and played it through Amarra and, like you, preferred the CD version over the SACD. Think I will stick to CDs, and hi-res downloads.

 

[jrebman]

Yeah, Jorma is one of my all-time fave guitarists.  I was scheduled to go to his guitar camp in the summer of 2000  if I'm remembering correctly), but that was the summer a bonehead doc damaged my transplanted kidney and the rest of that summer was a total waste.

Of course most any good guitarist in the style of Rev. Gary Davis is just somebody I have to listen to.

-- Jim

 

[Yoder]

I to have checked out his guitar camp but it is an expensive weekend, not to mention a drive through the mid-west.

I own a couple of Rev Gary Davis CDs, and have a song book with most of his notable songs in it. Just traded my thick bodied Gibson AJ, for a sweet thin bodied Martin 000-28EC. The Gibson really aggravated my shoulder so I played it very little, the Martin is a beauty and I have played it more in the last two months than I did the Gibson in 12 years.

Finally got all of my CDs re-burnt, and am now reorganizing the hd library. Somebody in the household dinked with my computer and I noticed that the jazz library was in m4a (sorry Doc.) Plus I have acquired about a 50 CDs since the last burn, and a lot of hi-res audio. Everything is now in AIFF. Once the Eros arrives and is built, then I have to rip about 300 LPs, hence any possible SACD conversion is a very low priority. I actually sold my Duet I (24/96) for the price I paid so that I can get an ADC that will rip at 24/192.

Anyway, what I want in a DAC?

ADC capabilities?

 

[earwaxxer]

like SACD uses the same licensing scheme as HDMI does. Ergo, SACD audio can only be played through analog outs. As soon as I change the output to my SPDIF or optical the player reverts to CD mode. Then it is in big print in the manual that "SACD can only be used with analog outs." Bad news is that one is reduced to using the DAC in the SACD player, but if one wanted a better DAC then there are always the high-end CD/SACD players available. So, I was unable to see if there is a much difference in down-sampling from 96 to 44. The good news is that I can still burn the Redbook CD to my hard drive.

Hey Yoder - Yep - a shame it is! Now you have outfits like HD tracks that rip SACD's and sell them as a high res. 'remaster'. Its BS. High res will stand on its own if its worth it. Audiophiles are picky bunch. We can find the chink in anyone's armor.