|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
142.0.195.166
In Reply to: RE: Yes! posted by Ivan303 on July 10, 2017 at 21:57:29
And it sounds great! ;-)
Follow Ups:
The original recording was DSD64. All other rates are either upsamples in the DSD domain or PCM transcodes. None will contain more musical information than the original DSD 64.
Upsampling and transcoding can (and usually do IMO) produce unwanted artefacts which may be very subtle but may neveretheless be audible. So the simplest answer is to keep to the original recording format. However this is not possible here because, as with the the majority of DSD recordings, it has had to be converted to PCM for editing and then back to DSD. The latter process does not necessarily mean that any artefacts from transcodimg are obviated.
So, as all of the formats offered are compromised and as there can be no additional musical information in the higher rate conversions, two other factors become important.
The first is hardware i.e the DAC and how good the digital filter or filters ( if you have a choice) are with the particular format. This will have more effect upon the resultant sound than the format itself. In fact as far as I can work out the only advantage to upsampling is the ability to place the pole of the filter further outside of the audioband ( which may also allow use of a gentler filter characteristic albeit with lower suppresion of aliases). As the filter seems to be the trump card this accordingly provides an unpredicable situation for recommending formats on a general basis.
The second factor is cost. As the DSD64 version is less costly than the others and as it was the original format unless I had experience that one of the other formats always offered better sound with my particular equipment I would save my money and go for the DSD64.
Chris, that does not mean that your DXD version does not sound great. It does, but with your equipment. No guarantee that it will be better or best elsewhere.
NB: Out of interest I spent a merry hour or so last month converting redbook files to both DSD 128 and DXD. The result when going back to the original redbook was similar to removing a stone from a shoe. So nice when it (the upsampling) stops.
Here are Jared's own words:The recording was originally digitized using the Grimm AD1, which operates at DSD64. The original session tracks were edited and rebalanced (which meant going through the mixer) in the only available format for that purpose; the Pyramix 352.8KHz/24bit PCM (DXD). Prior to the advent of direct digital delivery, the next step in the production process from 352.8KHz/24bit PCM would be the DSD64 edited master for SACD production. What we have done now is also make a direct conversion to DSD128 and DSD256 from that original DXD edited master, without going through any interim processing steps.
Those DXD to DSD conversions are not up-samplings, as they would be going from one PCM sampling rate to another, for they are different encoding systems. PCM is a digital value sample based system, and DSD is a digital bit density modulated system. Conversion from any PCM sample rate to any DSD bit rate system is a remodulation, not an up-sampling.
We feel there is an audio advantage to this process in using the original files so we give you the choice and you can decide.
Otherwise, I agree with what you said.
Your comments about upsampling vs. native resolution intrigued me. I have an Esoteric SACD player, and like several players these days the Esoteric offers multiple sampling/filtering options for RBCD playback to keep tweakers occupied. I played around with these settings right after I bought the machine, but ended up selecting DSD conversion as (seemingly) the best-sounding option for CDs. Are you implying that CDs would sound better using the straight 16/44 setting? I'll have to experiment to see how this plays out in my own system.
If I've misread or misinterpreted your post, my apologies.
" Are you implying that CDs would sound better using the straight 16/44 setting? ".
Not quite as I am saying that the digital filter characteristics will determine the "ranking" of any one format over another where sample rate conversion is involved. This may not be the case using native hi-rez formats as they will contain actual additional musical information.
If Esoteric provide various filters that may be used with straight 16/44.1 files it may be that use of one will sound better with this format than upsampling it to DSD with whatever choice of filters you have there. However as each player manufacturer can implement its own filter designs there is no common ground. So even if I believe that upsampling results in certain subtle sonic abberations the net result may still be preferable over redbook in a given system . It depends upon your particular set-up ( and preferences). If you are fortunate to have access to an asymmetric or apodising filter for redbook then I would expect the native resolution to be preferable. However given a typical 22Khz brick wall filter with its attendant pre-ringing, maybe not.
My view would also be that if possible where upsampling is considered necessary is to keep it in the same domain as the original and to use only multiples of the original sampling frequency. So ideally I would avoid transcoding from PCM to DSD and for 16/44.1 use 88.2 or 176.4 sample rates. Hopefully your machine also adds 8 digital zeros to provide 24 bit depth. Again no additional information is created but this can be an advantage with many digital processors where they are designed to process 24 bit words.
I can only suggest that if you have been using DSD upsampling for a long time without question it may be worthwhile switching back to 16/44.1 just to check that your decision to use DSD was correct - it may well be but it will depend on the filters and the upsampling algorithm that Esoteric have provided.
"None will contain more musical information than the original DSD 64."Sadly not. Sony never completed development and sale of a commercial mixing system for DSD.
DXD is as close as you're gonna get, at least in this case. For all intents an purposes it is the 'master' as it came off of the mixing board.
The thing is, Jared is being very honest about his process so you can choose your best strategy to get as close a possible to that original recording.
I just wish he would offer the DXD files as download options in the future.
I intend to download the Mahler 3 DXD at some point, soon as I figure out how to play it. ;-)
Edits: 07/11/17
The link is to the Channel Classics catalogue. I know that my vision is impaired but I can't see any reference to a DXD issue of the Mahler 3. Jared refers only to DSD 128 and 256 files transcoded from the DXD file.
I will remain sceptical as to any advantage of these over the DSD 64 version. Although I am always interested in the proof of the pudding I won't be buying the DSD64 and DSD128 versions to compare, mainly because I don't like Mahler much!
and only one has the DXD for sale?
Maybe. I suspect a record company cock up. Interestingly one site is the Channel Classics catalogue, the other is Native DSD which I think is owned by or otherwise related to Channel Classics but offers other labels.
Interestingly if you buy from the catalogue website (without a DXD offer) your order will be processed by Native DSD ( where the DXD file is available) as they are linked. I'll look for a handy nearby wall for some head banging.
I'll jump in soon as I am yet to figure out how to get DXD to play on iTunes (with Bit Perfect plug in). The Bit Perfect plug-in for iTunes does not support FLAC so I have to convert to something Bit Perfect will play while maintaining the sample rate of DXD.
All very complicated for an rapidly aging brain.
. . . although, come to think of it, I haven't actually tried playing multichannel files while Audirvana was still operating in "compatibility" mode. Maybe I should give it a try and report back.
back when I dawdled with Mac.
. . . because iTunes has a WAY BETTER user interface (IMHO) than, for instance, HQ Player does. Too bad about all of iTunes' other limitations. As my library of multi-channel files continues to grow, I find myself (of necessity!) using iTunes less and less.
I did try it a couple of times, almost by default on a couple of machines. However, I found the UI totally counter-intuitive and the format/program constraints unacceptable for classical music.
That was a long time ago, so things might be a bit different now but I use programs that are much more configurable and which have little or no constraints on how I use them or on what files I can play (including the rare AIFF).
I know of nothing about it to tempt me.
iTunes will not allow a multi-channel aiff file to be added to its database, so the only way that Audirvana can play multi-channel files is via its own database. (So compatibility mode needs to be switched off in order to play the multi-channel file.)
prefer the SQ of iTunes/Bit Perfect.
And stereo headphones so I don't care.
I will be downloading the DXD Stereo version anyway.
It never occurred to me that there actually were 2 channel DXDs! Duh.
most of us were born with only two ears. You guys seem to have FIVE or even MORE! ;-)
And yes, navieDSD will sell me the DXD file in multi or stereo but from their web site:
DXD = 32/352.8k PCM - 8X Redbook [file type is .wav and sometimes .flac]
I can import .wav into iTunes but not .flac. :-(
AFAIR, if you import a 24-bit WAV file into iTunes, iTunes will simply truncate the last 8 bits, resulting in a 16-bit file which is not optimized for 16 bits - unless something has changed in the last few years.
OK, maybe 18 to 20 bits, but no more.
via the USB input (Amanero Combo384 Module).
Not sure I get the benefit of the 24 bit depth but who needs to hear 'Brownian' motion? :-)
As the DXD file is FLAC I will need something to convert it with.
Any suggestion for freeware and is there any other CODEC other than FLAC that will store 24/352.8?
I used to use a product called "Max" - although never for DXD. It also ran afoul for me when converting multichannel Chandos flacs to aiffs. (Two-channel flacs were always OK. And even with the multi-channel problems I had - with the wrong signals ending up in the wrong channels in the converted files - ONLY Chandos flac files were affected - the conversions from BIS flacs were always fine - go figure!)
I also have a conversion app called XLD, but I haven't used it yet and I don't know what its capabilities are. I also think that dbPoweramp might work too.
I don't understand paying more money for a DXD file and then converting it to something else in order to play it.
Alan
24/96 flacs converted to 24/96 aiffs.
I only have a couple of DXD files (so far anyway), and I play them as flacs (i.e., as downloaded).
Or are you talking about ivan? I'll let him answer for himself.
No, It is my understanding that Channel masters in dxd and then will sell you different versions derived from the DXD master. Seems overly complicated
Alan
in Stereo or Multi-Channel.
Sadly, this is NOT the case with most other Channel Classics releases. :-(
Some listeners have personal preferences for one format or another, and are willing to argue about it for years! ;-)
I actually applaud Jared for trying to accommodate these different types of listeners in this release.
To me it looks like these various formats are just a lot of technical hype for marketing purposes. It raises the price of distribution, and ultimately the price to the consumer. I don't believe you can go from one format to another without losing something! Whether it's audible or not, and to whom, might be the issue.
If they wanted to give us the very best, the "mix" would have been in the original miking as was done on the so called "non-commercial" DSD256 version written of above. If DSD64 is what they were working in (and I guess it could have been DSD256) that could have been the principal release. In the glory days of vinyl, were not many recordings made that way? (Too many mikes these days?)
So then they convert to 352.8KHz/24bit PCM (DXD). Question: If they felt they eventually need to go to the Pyramix, why did they not use 352.8KHz/24bit PCM (DXD) for the original live recording? Then they could have done the "required" mixdown with no conversion and that would represent the best that it would be possible to offer. [But they would lose the marketing cachet of having recorded in DSD v. the "old fashioned" PCM.]
But having converted to 352.8KHz/24bit PCM (DXD), how could the fidelity improve by doing yet another conversion? I would have to conclude that 352.8KHz/24bit PCM (DXD) is the best they can now make available.
Given the sophistication of the playback equipment likely used by those considering either the 352.8KHz/24bit PCM (DXD) or one of the DSD versions, is it likely that a DSD version will play back more accurately than the DXD? Is it not accuracy that is the goal? Or is it meeting the personal preference of an end user unfamiliar with the original recording experience?
Finally I might express the view (which a multi-channel listener could feel free to correct) that those of use who listen in just two channels may feel a greater need for accurate aural cues for the space and placement of instruments.
In this case, I would conclude that too, but, as I mentioned, not all listeners feel the same way.
in the recording industry is that Jared actually reveals how he makes the sausage.
You can record in DSD but you can't mix in DSD so you convert the DSD to the very best PCM format available which I trust Jared uses in the Pyramix.
On can argue all day long about formats, PCM vs. DSD and bit depths and sample rates but the truth of the matter is, the problems are largely DECODING issues in the various DACs and NOT in the information contained in the various formats.
In other words, if the recording engineer got the recording part right and it wasn't screwed up in the mixing board, it's on YOU to figure out the best way to play it back.
I've heard Jared's recordings played back on Gigabuck gear and can safely say he is producing some of the best recordings being produced today.
"In other words, if the recording engineer got the recording part right and it wasn't screwed up in the mixing board, it's on YOU to figure out the best way to play it back."Interesting, and just how are we to figure it out, aside from theoretically, to wit, the DXD files.
I appreciate the samples and I downloaded 2 of them, the DSD256 and the DXD. Problem is that the DSD file has little more than adjusting the bass drum sticks. The DXD has more. Perhaps I should try others. But these do not represent the miking and mix of the final product.
Better still if Jarad would distribute a six minute sample of the final mix that he thought best illustrated his work, and in the various formats.
Edits: 07/12/17
IMNSHO, it's gotta be user hardware dependent.
There are so many choices in how to decode a digital file that I doubt anyone would ever agree.
I use a multi-bit ladder DAC that likely has a real bit depth of about 18 bits or so. It's based on 8 each PCM1704's and would by many be considered obsolete as TI doesn't sell the chips anymore.
Others have DSD DACs fully capable of decoding DSD-512, some use software to up-sample to that rate, some DSD DACs up-sample internally to even higher sample rates and even deeper bit depths and on and on and on.
Who is to say what's best for you?
I use iTunes with a $10 Bit Perfect plug-in so WTFDIK
Hi Mel,
You're correct, the free samples on NativeDSD are not the micing and mix of the released album, but are a single take snippet using different mics, recorder, and placement. But the samples do provide accurate comparitive examples of the different formats (PCM/DXD and DSD rates) as how they will play and sound on different playback systems.
Your's is a reasonable request of providing a six or so minute sample of the released Channel Mahler 3 recording in the various delivery formats/rates, so we'll see how best to distribute these.
Thanks for your suggestion!
I spoke with Jared, and he'll add the Mahler 3 5th Movement to the NativeDSD Just Listen label page, in all the deliverable forms next week. This movement is particularly useful as a format comparison, as it is very rich sonically containing not only the orchestra, but the women's and children's chorus as well as the alto Gerhild Romberger.I'll keep you all informed.
Tom
Edits: 07/13/17
Due to vacation schedules it was not possible to place the Mahler 3 5th movement on NativeDSD this weekend. Until we can get it placed, if anyone wanting the movement earlier would mail me at the Audio Asylum tailspn address with the format and speed desired, I'll E-mail a We Transfer link to it.
Edits: 07/22/17 07/22/17
and iTunes with the Bit Perfect plug-in won't play FLAC. The nativeDSD site does not specify which file type the download.
DXD is 32/352.8k PCM and it doesn't necessarily have to be a .flac file, could just a easily be a .wav file and it would remain a DXD file and would still be 32/352.8k PCM
See link below:
Or, it could even be a 24/352.8 aiff file. ;-)
if there is such a thing?
My freeware conversion only goes to 192 :-(
Do all these different formats really sound significantly different or are they small. For me it is still the original quality of the recording that matters. For example I really don't care for the sound of the Channel Classics recordings. The don't have the immediecy of for instance the old Decca, RCA and Mercury recordings. I think they are using to many mikes.Read a review of one of there recordings which said they used 17 mikes. Now it is really up to the mixer to create the sound they are looking for.
Alan
if they are truly lossless.The real issue here is with the playback software you use, will it even play it, how much CPU time it takes to convert it to play, how much space the particular lossless CODEC takes on the hard drive, how much it can be compressed for streaming or downloading, etc.
Edits: 07/12/17
They're very small - as I always say, it's like arguing about how many angels can dance on the head of a pin. (FLAC files are losslessly compressed, whereas AIFF and WAV are usually uncompressed.) And you're right that it is still the original recording quality which matters most, regardless of the particular incarnation.
Yeah, I like the old Decca, RCA and Mercury (and EMI!) recordings too. But OTOH, I believe that great recordings can be made with different microphoning philosophies, and, certainly (or at least IMHO), Channel Classics invariably gives us outstanding examples of their recording philosophy with each new release - really state of the art for their particular approach.
We can't possibly be defining state of the art dependent on what recording method is used? Can we? State of the art should be the absolute best, period!
Alan
You say that "State of the art should be the absolute best, period". And yet you also ask "We can't possibly be defining state of the art dependent on what recording method is used? Can we?". If we're talking about absolute best, then I guess we'd have to say that, yes, one and only one recording method must be superior. (Or, one and only one microphone set-up - or whatever.) I don't believe this myself, so I guess my definition of "state of the art" is not as narrow as yours?
I don't really disagree with you. There are a lot of great recordings out there. Having listened to tons of recordings and also recorded both multi mike and the 3 mike Decca tree method I think there are more great sounding recordings using 3 mikes than any other system. The move away from 3 mikes as the technology changed was the famous statement Will fix it in the Mix. Works sometimes and also fails sometimes. As far as multi mike recordings go I think Reference Recordings have done the best work.
Alan
or buy what you can play.
which I understand is a fairly common occurrence or there wouldn't be all of that conversion software available.
But the conversion freeware program I use won't go past 24/192 and I'm not certain what file type I should be using.
Bit Perfect WILL handle up to and including 384kHz sample rates, all I have to do is figure out how to get it into iTunes in a format that Bit Perfect can play it.
iTunes is for folks with two ears, not five. :-)
lossy 256k content. :)
will play just about any sample rate you choose, up to and including DXD. But not from a FLAC file.
But I also have an Audirvana plug-in that plays nice with iTunes
.
of being 'with Kal'. :-)
his previous comments of course. :)
I do run an Acoustat based MC/HT system although the only stats are the mains.
I'm trying to sell my original non-PX cores to a guy with three U-1s and spare backplates who wants to create a five channel SL system. Not quite as much panel area as Ray Kimber's ten Prostat 922 multi-channel system, but still a bunch!
I've got too many coffin sized boxes in storage. :)
Worth the drive to Denver (I lived in Utah just down the road from Kimber Kable during many of those years).
If you're gonna do multi-channel, I guess ya gotta do it right!
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: