Do we need hi res files?

[Scroobius]

A couple of years ago I was listening to music on my audio note system (AN DAC & monoblock SET amps) with XXHighEnd at the front end. At that time in my mind the best recordings I had were iTrax 24 96 HiRes files. There is no doubt (at least as far as I know) as to the quality of these digital recordings both the recording techniques and the technology used to produce the 24 96 files. There is no "dodgy" upsampling with these files.

But now my system is completely changed and the sound quality is in a totally different league. So how do those 24 96 files sound now with a NOSUSB based system? Well I had not played any for a while and I have to say I am distinctly underwhelmed and I do not feel that I need to use hi res files to show off my system any more.

Maybe, just maybe, at their best they are as good as some of the red book material. But generally I am not sure if they are any better. To show off my system now I would use some of the best red books I have, and some of those are from the 50's and 60's.

So as Peter has mentioned many times before I would now agree that it has to be questionable if HiRes really is the way to go. Now for the first time I can really understand with my own ears what Peter is talking about.

Well that is the way it sounds to me anyway for what it is worth.

P

[Flecko]

There was a discussions about that topic. As far as I remember noone felt to need hires. What we need are good recordings that are not messed up during mastering. A further improvement would be a medium that allows a secure read out of the data. A data format that can be reproduced in a secure way without ringing and bandwidth problems. I saw an impuls response of an active speaker with optimization of the impuls response. The peak was very "exact" in time of all three drivers but it was full of pre and after ringing. I cannot beleave that this is the way to go. Without hearing the NOS1, I think this is one important part of the improvements to be done. Hires is more a marketing joke to me. This "joke" comes with digital remasters which are for sure reduced in soundquality due to dynamic compression. But I do not think I am saying anything new to you. Anyway. We can write those things to stone and keep them as the ultimate laws of digital audio at 2011.

[Robert]

You pose an interesting question. Of all I've downloaded only one stands out to me Miles Davis-Tutu from HD Soundtracks USA in 24/96. It definitely is a big improvement over the CD and vinyl both of which I own plus I heard this live when Miles did his world tour to promote this album. One of the best concerts ever for me.

Unfortunately there is so little music that I like in hi-res which has always been a problem with so called audiophile recordings. Friends that have these tend to use them to demo or show off there systems, feel the bass, imaging etc. They are not particularly interested in the music only as a tool.

I agree that some of the older recordings in red book are excellent played back but I also love the music warts and all.

I'm very disappointed for example Pink Floyd recently remastered all there albums at 24/192 but publicly these were released on 16/44 CD's or if you bought the boxed set you got a blu ray with 24/96 on it which Chris Connaker(Computer Audiophile) recently ripped to hard drive. Why are these not available for downloads at a price I'd be happy to pay. The 24/192 must sound amazing.

Bjork's latest album Biophilia download is available from her site in WAVE wow! but only 16/44. Why not offer 24/96?

Part of the problem is record companies still have not got there head around this whole download situation and we as audiophiles are still a very small part of the market less than 1%. Lets face it MP3's have replaced CD's and the urgency to provide high res is minimal. 

Recording studios are another whole issue. This is a small piece from Barry Ober The Sound Doctor JL Audio who has extensive experience with studios:

THE RECORDING PROCESS

On top of all the previous variables we have all the issues and errors inherent in the recording process. It is simply laughable (and pathetic) when I read the magazine articles where the "soundstage" of a rock recording is "palpable". Sorry, but every modern rock recording made in the last 40 years is composed of a series of panned mono signals that have absolutely no "depth". They are each separately sent to an echo/reverb device, the returns of which are usually (but not always) panned full left and right. The summation of all the L-R panning placement and the summation of all the reverb returns fools you into thinking there is a "soundstage". Sadly, precious few recordings are made with any regard to true stereo or binaural sound in anything resembling a true form; even better classical recordings of large orchestras have morphed into combinations of stereo miking and "some" local more-nearfield mono miking added to the mix to achieve whatever the producer determines is a suitable balance, perhaps between a soloist and the rest of the orchestra. Yes, there are precious few companies who do pay attention to this; AIX records is one. But to think that any modern, commercial pop recording mix has any true acoustic space is, for the most part, sadly mistaken.

This is a piece from connecting subwoofers to systems http://www.soundoctor.com/whitepapers/subs.htm  but this bit on recording I find depressing reading.

Its still early days though and hopefully more high res stuff will become available. I personally think the artists need to take over and offer them directly. Potentially 24/96/192 is amazing and a big step up from 16/44.

[pedal]

Yes, of course we do.
When done right, hi-rez simply sounds (slightly) better.

BTW: Thanks to the ARC Prediction function, the combination of XX/NOS1 allmost closes the gap between CD and higher sampling frequencies, when it comes to natural treble.
With my previous Buffalo II DAC I heard a "clearly better" treble when comparing CD and hi-rez download of same master.
-Repeating the same comparison with my NOS1 USB, they are very, very close.

But hi-rez is not only a matter of ultra sonic freqencies.
The 24bit resolution provides better accuracy for all frequencies. The sound is more "fluid" and "analog". There is a certain "naturalness" with (good) 24/192 that makes even the best CDs sound slightly "reproduced". There is a "16bit signature" to the sound.

It's a common misconception that more bits means more dynamic range only. -That the 16bit resolution is good enough for all signals within the the 96dB range of red book CD.

My guess is that whats going on below -96dB is the least interesting benefit of increased bit depth. -The improved accuracy within the 96dB window is more important. What I hear as a kind of smoother rendering of micro dynamics. You hear it on acoustic bass too.

[praphan]

I fully agreed with Robocop.

1. Hires materials from labels like Chesky, RR do make sonic difference. Not only me but my wife and my children can tell the difference between similar track recorded in red book and 24/192 resolution. The most noticeable difference is the wider space of sound stage and better sense of location of the instruments within "that" stage. Hirez materials offer higher level of musicality and fluidity. This is why firm like Chesky is very stringent in marketing their hires albums. Chesky only ships its hires albums in the USA.

2. Unfortunately, audiophile record labels cannot afford to sign contract with grammy award winning artists. So pretty rare to enjoy world class music with hires sonic quality.  

3. We are all aware of the fact that many hires albums offered on the download market are just "fake" upsampling version from red book. We should not waste money on these.

Praphan

[Scroobius]

I have to say that I used to hold exactly the same view i.e. that hi res provided much better sound quality - after all in my old Audio Note system hires was clearly better - but then came NOS1USB in my new system and it completely changed my view. Red book now offers astonishing sound quality through NOS1USB.  So now I am not sure if hi res offers any improvement. BUT I really am going to have do more listening tests but it is difficult because it is virtually impossible to get hold of files in both formats that can be compared on an equal basis.

One area that surprises me is that always before I would point to dynamic range of hir res material as being superior (and that is what I thought my ears were hearing) but now that difference appears to have disappeared (but as I say I really must do more listening tests). But why? NOS1USB for sure allows the excellent upsampling ability of XX to work its magic but it cannot do anything about the bit depth of the original material. I always thought that 24bits had to be better than 16bits. Maybe 16bits really is enough - after all it is capable of 96db dynamic range and I have just read that "The dynamic range of music as normally perceived in a concert hall doesn't exceed 80 dB". So I have to ask myself what did I think I was hearing before?.

All I do know is that now if I want to show my system off I will use some of the best red books. Those early 50's to 60's really are some of the best recordings - in my old system they did not sound anything special at all.

NOS1USB has thrown a great big spanner in my previously held hi res "belief's" I really must do some more listening tests though - call I can say is that now I am somewhat confused on the matter (my wife would say "nothing new there then").

[Scroobius]

Good point by Pedal 24 bits gives more samples and therefore more accurate sample values (ie less rounding off) so there should be less distortion (i.e. less requirement for dither). I agree it really should be better. 

[pedal]

3. We are all aware of the fact that many hires albums offered on the download market are just "fake" upsampling version from red book. We should not waste money on these.

Yes, unfortunately there are some stinkers among the hi-rez titles. But we shouldn't focus too much on the bad ones. With time, HD Tracks and the rest of the music industry will improve their hi-rez quality. The major artists will join too.

But don't forget that major artists like Bruce Springsteens mix and master their albums monitoring through a ghettoblaster.    Transfering these rubbish mastertapes to hi-rez doesnt make much sense...

[PeterSt]

So - Anyone with an example of Hires better than Redbook ? And to keep in mind : you are not supposed to come up with a different mix or anything.

We should all (ok, those having both the versions) be able to judge and tell about the differences.

Curiously waiting,
Peter

[Scroobius]

Good point Peter but the only decent hi res tracks I have are not available on red book so I cant compare. They are the iTrax recordings and whilst they are very good well I really don't know if they are inherently any better then red book. It still seems that the overwhelmingly the most important thing is to have a great recording quality in the first place.

[PeterSt]

Well, that is what I meant of course;
If nobody is going to come up with a few examples (let alone only one), the same "nobody" is not allowed to determine HiRes sounds better.

All in the context of all the many examples which exist which *are* from the same versions and which all totally fail.

The only thing which seems possible is having that random Hires album which perceivedbly sounds good, and compare that to any other random Redbook, of which I claim the latter sounds better. So, play 20 Redbooks and next play any HiRes of your likings, and the latter sounds flat, dull, has strange layer effects (actually lacking layers).
But it really needs something like the NOS1 in the first place, because otherwise you will be comparing apples with oranges;  not because the NOS1 shows HiRes better or worse than a random other DAC, but because the other DAC is no representative for Redbook.

And to keep in mind : originally everybody says that Hires sounds better (not me, ever). Until they saw the graphs coming from it, and the same everybody was instantly cured. Not much for realism, honesty, good ears or a good system, but the graphs *do* show the truth.
All 'n all be careful with your examples of good HiRes in general, because 10 to 1 *I* will show you how poor it "graphs". Unless you did that yourself of course, and hold your at-first-example back because of it.

And oh, you possibly can come up with OK albums which were recently recorded in HiRes. But it really makes no sense because within a few days you will have heard them all and there it stops. And then *still* it will be about the recording quality in general, the "performance", the music, the artist - and within seconds e.g. I will show you a far better sounding other album in Redbook.

Btw, that Bob James from the other topic is such an example of the better Redbook. It's not easily outclassed by something else, just because of the recording. So see ? that handful of modern HiRes juat hasn't got a chance. The recording will be more poor etc. By accident of course, but this is how statistics work.

Peter

[pedal]

So - Anyone with an example of Hires better than Redbook ?

Doing this in a scien'tistic way, I think there are 2 ways:

1. I have some CDs from Chesky Records which I also later bought as hi-rez downloads. I have compared them in the past, and they sound very similar, except for the tiny hi-rez difference (smoothness). I will double check this to see if we truly compare "apples with apples". Stay tuned.

2. The second solution is much easier: Take any well recorded hi-rez album in 24/192 and convert it to 16/44 with a suitable software. (This is what they do in the masterstudio anyway). Then listen and compare.

PS: Peter - I am surprised that you dont recognize the benefit of 24Bit recording and reply, opposed to 16Bit. 

[PeterSt]

2. The second solution is much easier: Take any well recorded hi-rez album in 24/192 and convert it to 16/44 with a suitable software.

That is what you think ! You may have no idea how many "settings" even one such a software contains, and what you can do all over wrong. No, not for me.

PS: Peter - I am surprised that you dont recognize the benefit of 24Bit recording and reply, opposed to 16Bit.

Maybe if I take the recording ...
So, I don't think I said I don't recognize the benefits. They should just be there, and they will be. But I listen. And then it doesn't work out. Remember, I am talking about the larger gamma of available HiRes, and not today's handful of recordings. This more old stuff - or even today's remasters like Waltz for Debby ... they all totally s*ck. The reasons vary very much (really dozens) but there's always something and it can always be proven by measurement.

Regards,
Peter (listening to NHOP's Scandinavian Wood ... why would I go searching for anything better ?)

[Flecko]

Anyone with an example of Hires better than Redbook

I am on the "we do not need hires" position but I have an example:
Antonìn Dvořák - Symphony no.9 Op.95 "From the New World" - N. Harnoncourt released on telarc
DVD-Audio is slightly better than the cd.

[pedal]

...So, I don't think I said I don't recognize the benefits. They should just be there, and they will be. But I listen. And then it doesn't work out. Remember, I am talking about the larger gamma of available HiRes, and not today's handful of recordings.

Hmmm, seems we are viewing hi-rez from different angles. I am stating the obvious; The original SQ of a mastertape is better preserved in 24/192 than in 16/44. For this reason I want consumers to support hi-rez formats, so the music industry will be encouraged to increase their hi-rez output. Although there are some stinkers, I think the hi-rez catalogue will improve both in quality and choice of titles. Sooner or later you will get a flat hi-rez transfer of the NHOP mastertapes too.

I guess the ratio between hi-rez titles and CD is 1:10000. Of course, right now it’s easier to handpick super sounding CDs, simply because there are so many to choose from. And if, additionally, you want to keep modern hi-rez recordings out of the picture, then the “match terms” becomes rather unfair.

All in all, I do applaud the arrival and growth of hi-rez music

This more old stuff - or even today's remasters like Waltz for Debby ... they all totally s*ck.

Really? All of the old stuff sucks? -I dont think so.
The oldest hi-rez album I have is probably the 24/96 version of Miles Davis - Relaxin With The Miles Davis Q. Track 2: You're My Everything. It's a recording from 1956.

You have it too, I think. Give it a listen and tell me if it sucks 

[Scroobius]

Hey Pedal,

2. The second solution is much easier: Take any well recorded hi-rez album in 24/192 and convert it to 16/44 with a suitable software. (This is what they do in the masterstudio anyway). Then listen and compare.

If you down sample in the way you suggest then you HAVE TO apply a brick wall filter so that images (per Nyquist) do not appear in the audible band in the 16/44 file. That is potentially a VERY VERY bad thing to do to the sound quality and means it would never be possible to do a meaningful comparison.

As an aside it is deeply sad that recording engineers do not appear to understand some of the basics and that if you upsample or downsample there are potentially big prices to be paid in terms of sound quality. Well in terms of upsampling Peter has shown it is possible to do a good job - but how many others do? none that I know of. But in the case of down sampling well it is the work of the devil and should never be done.

The original SQ of a mastertape is better preserved in 24/192 than in 16/44

Well technically for sure that may seem to be correct - but does it actually work in practice? - I do not know the answer to that question and at the end of the day there is only one way I am going to get an answer and that is to use my ears - I certainly would not take the word of anyone in the recording industry.

P

[Flecko]

If you down sample in the way you suggest then you HAVE TO apply a brick wall filter so that images (per Nyquist) do not appear in the audible band in the 16/44 file. That is potentially a VERY VERY bad thing to do to the sound quality and means it would never be possible to do a meaningful comparison.

It is maybe affecting the sound but it is by far not the worst thing that can happen. Using a bad sounding reverb, overdooing it with equalizers or compressors can realy destroy the sound. And that is what is done in the first place. We were lucky if we could realy complain about downsampling. Also there is an advantage of recording in an higher sample rate because of the headroom that is needed to mix the musik.

Well in terms of upsampling Peter has shown it is possible to do a good job - but how many others do?

I know this is some fundamental thing and it is likely that not much people agree but the best dacs I have heard are "normal" upsampling dacs. Not that I heard the best dacs in the world but at least they are not that bad even if upsampling is done in the "normal" way. So, what I want to say is, that upsampling or downsampling are not what we should argue about. The bad guys are the overmixing and overmastering tone engineers (and all the mechanisms that lead to their incompetence). This affects the sound WAY more. And to change that it "just" needs some education and no technical revolution. Every thing is there to create a good sound from pocket money recording equipment. It just has to be done in a right and sensible way.

[PeterSt]

The bad guys are the overmixing and overmastering tone engineers (and all the mechanisms that lead to their incompetence). This affects the sound WAY more.

This is true. And sadly, this is the very first thing what "is applied" when HiRes is created out of originally good stuff. And there it fails. Completely.

Really? All of the old stuff sucks? -I dont think so.

Pedal I did not say that, and certainly did not imply that. You know in what realm I judge and write about it.
HOLY sh*t (by Bill Evans).
So please read better (or otherwise I must write better haha).

I am talking about the larger gamma of available HiRes, and not today's handful of recordings. This more old stuff - or even today's remasters like Waltz for Debby ... they all totally s*ck.

This is what I said, and it plainly tells about the older HiRes (early 2000's with nothing in between then and a few years back), all s*ck. Proove me otherwise.

More in the next post.

[pedal]

This discussion is all Déjà vu. Let’s put the hi-rez controversy in perspective, and paraphrase the topic of this thread:

"DO WE NEED STEREO?"

-Many listeners were dissatisfied with the first generation of stereo recordings in the 60s.
The first records sounded very annoying, with the vocals pan potted to the extreme left and the instruments to the extreme right channel.

Listeners felt that the established technology (mono) sounded better.
Even the musicians preferred mono. The Beatles recorded all albums in mono up until Abbey Road in 1969. (Stereo mixes was an afterthought, post-produced in the studio, mainly for the US marked).     

Because the recording industry hadn’t adopted the new technology, the first examples of the art were not so successful. At least not for pop/rock. (Early jazz/classical fared better).

--------------

There is nothing “wrong” with hi-rez, technology wise.
What’s wrong is the implementation of the technology.

But it will improve. Hi-rez is the future!


PS: Peter, did you listen to the Miles Davis track #2 from 1956? It's in glorious mono! 

[PeterSt]

This more old stuff - or even today's remasters like Waltz for Debby ... they all totally s*ck.

Really? All of the old stuff sucks? -I dont think so.
The oldest hi-rez album I have is probably the 24/96 version of Miles Davis - Relaxin With The Miles Davis Q. Track 2: You're My Everything. It's a recording from 1956.

You have it too, I think. Give it a listen and tell me if it sucks 

What did I just say ? they all s*ck. This one too. Or especially, because of the poor distorted sound and all the anomalies in it;
It :

- Is no Hires at all (hard cut at 22.05);
- Is clipping in each track (not severely, but it does -> fine engineering);
- It does not contain any info under 100Hz, while there's a severe anomaly at 100HZ (should be some kind of hum);
- Sounds as flat as a pancake.

So, it fails all over, and this is what you put forward as an example-of ? I don't get it.

It doesn't matter with which one you come up, because here too, statistics proove that I will judge/write like this about any HiRes (older !) album anyone comes up with. So to be clear : not necessarily about today's.

I hear you say : yea, but this is 1956 ! So what ? read that Waltz for Debby topic I just pointed you at. Same story, although 5 years newer for that particular album, though remember I went back to 1958 I think, and all is just the best. Until, until it is remastered (which is different from a first transfer to CD), may it be just that remaster or a remaster for HiRes (which usually was for multichannel; this Miles Davis too if I see it right in the graphs). Really, I don't even need to look when it's 24/96 (24/192 is another story).

For this reason I want consumers to support hi-rez formats, so the music industry will be encouraged to increase their hi-rez output.

You sound like plugging something ?
It won't happen because it is too late. No digital sound engineers around to do it, and otherwise the tapes have worn out if they can be found at all.

Sooner or later you will get a flat hi-rez transfer of the NHOP mastertapes too.

... which will be the moment I don't play it.

Sadly I don't have the "original" CD album of this Miles Davis, but I can imagine I can dig it up somewhere. Otoh, it costs too much time to obtain the proper one. Like with Waltz for Debby; this example took me around a day to work out which version I have, which is a best one. All others are not at all (just the Redbook versions/remasters) and *still* I couldn't point you to this version. So, never mind.

The latter implies the danger of liking "HiRes" better than that Redbook version you coincidentally have, because it will just be another remaster. So, any remaster sounds completely different, and the one is worse than the other. This means that comparing already is a difficult task because you'd first have to *know* that you have the most original one. Like one which went to CD in the early 80's or so. When people didn't understand how to molest anyway.

I have quite some emails from people like you, who will never give up on their IDEAS of HiRes. It sounds better such and it is better so. A bit of a pitty if such an example isn't HiRes in the first place, don't you think ? This one, by accident, is in the open. Not really my fault, but it is a nice example.


My own example of good HiRes usually is about Beck's Sea Change (24/88). Technically good, but at the 4th track you will be sleeping. And then to think I never compared it with a Redbook version; it's just technically good which is audible in the absolute sense.
Also technically good are a few Alice Cooper albums in 24/192. Here though, it's the problem that it flaws for the recordings themselves (just poor sound, and not my music either). There isn't much more around.

Possibly the DSD originals bring something for the better. But this has its own problems, and the only confidence we can have is that it won't be manipulated (mind you, when original indeed, and which sure will exist).


Peter

[pedal]

What did I just say ? they all s*ck. This one too. Or especially, because of the poor distorted sound and all the anomalies in it;

It is interesting how we perceive sound quality in different ways.

I love the track You're My Everything - Miles Davis Quintet/Relaxin, album downloaded from HD Tracks at the modest price of $20 in 24/96. It was remastered by Rudy himself in 2005.

The song opens with studio chatter between Miles and Rudy Van Gelder, literally transporting the listener into the studio, witnessing Miles Davis in his prime. Poor Rudy only had a couple of microphones. The 2 horn blowers Miles/Coltrane shares one, and the piano and the rhythm section sharing the rest.
Of course there is lack of bass and treble. It is recorded in 1956, after all. But when I listen to “historical” recordings like this, I tend to filter out the anomalies. The muted trumpet is remarkable present and “in the room”, even more than on some modern recordings.

So, it fails all over, and this is what you put forward as an example-of ? I don't get it.

It was just an example of splendid preservation of a 55 year old 2-track mono recording.

Sadly I don't have the "original" CD album of this Miles Davis, but I can imagine I can dig it up somewhere. Otoh, it costs too much time to obtain the proper one.

It can probably be found here somewhere, 2nd hand as low as $5.-.

But according to Rudy Van Gelder, buying the original CD is a waste of time. Contrary to your believe, he disapproves about the first generation CD from this particular session, due to bad sound. You can read the interesting interview here: in-conversation-with-rudy-van-gelder

BTW: I brought the 24/96 download to a friend. (You know who). He has €50,000,- vinyl rig. The 24/96 was much better than his 180g audiophile reissue. The vinyl was muted and dull in comparison. Then he pulled out a second copy. -A very old pressing. Apart from lots of scratches and surface noise, it sounded much better than the reissue, and more or less equal to the download (through a Buffalo II). Through my NOS1 the 24/96 probably sounds even better than the old vinyl pressing.

It doesn't matter with which one you come up, because here too, statistics proove that I will judge/write like this about any HiRes (older !) album anyone comes up with.

Never say never! You should try ELP/Brain Salad Surgery, then. The last track #9 Lucky Man on the 24/96 DVD-A. It was originally recorded in 1970, included here as a bonus track. Sounds fresh like a daisy, 41 years later. 

The latter implies the danger of liking "HiRes" better than that Redbook version you coincidentally have, because it will just be another remaster. So, any remaster sounds completely different, and the one is worse than the other. This means that comparing already is a difficult task because you'd first have to *know* that you have the most original one. Like one which went to CD in the early 80's or so. When people didn't understand how to molest anyway.

Oh, I have experienced lot’s of bad hi-rez albums. After purchasing hundreds I have paid dearly for it. But in spite of some stinkers, I am very fond of the good ones. And I applaud the increasing number of new hi-rez titles arriving every month. Old and new music. This market is certainly growing.

[CoenP]

Imho the original post of this thread was all about the redbook capabilities of the NOS USB.

I can now testify that the discussion hires or redbook is rather academic in this perspective. Some of my better redbooks sound finer that any hires recording ever did on my set.

Furthermore, technically I am not convinced that more DR and more ultrasone information make any sense. This is not wrt our auditory system but rather the limitations of the recording and reproduction technologies. Potentially the increase in linearity (needed to catch a signal in >20 bits) is the most prominent reason that hires can sound better.

regards, Coen