After Rossini → Rossini apex upgrade, I wonder what would be the best way to upgrade the internal network renderer module. Does dCS provide the optional network renderer module for upgrade? If not, how much would internal renderer module cost, so that I can clearly know how much I should invest for the 3rd party streamer (taking into acount for external case expenses etc.)
There is no upgrade available to the network board AFAIK.
It is not possible to provide any information on costs.
One issue is that the board is made by a company that also supplies other audio manufactures. So you could be buying effectively he same thing albeit in another box, perhaps not as well implemented. Of course you might get support for DSD 256 although that is currently a bit of a unicorn format.
You will need to compare carefully IMO, preferably in your own system. BTW, you may be more likely to find an improvement by looking the source that supplies Rossini in the first place. Several of us have found, storage and a switch or router that improves upon standard IT units effective. However I cannot say exactly why and regular readers will know I consider it efficacious but irrational 
.
Respectfully, it is not a “unicorn” format. It’s a growing HD format, and that’s why every single high-end DAC maker in the world supports it, including dCS with the Varèse.
O.K. Niche.
Niche indeed. I looked at a few DSD sites and from what I see 99% of the releases are either Classical or Jazz. Those are the two lowest selling genre of music.
Yep spot on. I live on a diet of mostly classical music and i care little for DSD 256. I only have a handful of recordings - some are very nice, but hardly worth the effort. Others may argue differently but the fact is that dsd 256 represents a tiny fraction of releases even in the classical world
Regardless of how widespread quad DSD recordings are or aren’t the issue I have found is the quality of the performances. We’re not talking about the Berliner Philharmoniker or Murray Perahia here. Some of them have been so bad that I couldn’t even listen all the way through. I’d far rather listen to a red book recording of a good performance.
Just noting again that our ever growing collection of demo tracks for Varèse launches, chosen by about five of us and covering a variety of genres is still over 90% red book.
Aurender n30,n20, n200 all of which I used with bartok sounded more clear. When I changed to Sonja2.2i, the sound was too harsh soI sold all the aurender gear. I am thinking of “tube” sound streamer but they are difficuly to find unless I go for tube dac/streamer like Lapizator…
By the way. Briscati M5 had a unique dynamic sound unlike most streamers.
1 Like
Just a couple of number to think about:
Number of CD albums released ( all types and periods) 2,700,000 ( CDDB)
Number of DSD256 albums released 456 ( Native DSD)
IThese are numbers today’s computer search came up with.
And most DSD on nativeDSD is 256 DSD captured then transcoded to pcm for editing then back to DSD for commercial retail. I think there only 2 or 3 labels that release pure DSD recordings - basically live mic capture with minimal edits.
1 Like
This is correct, but it’s a little more complicated than that.
A native DSD256 recording does needed to be converted to DXD for editing on the Pyramix workstation software, and then reconverted back to DSD256. This is why Merging Technologies (the Swiss maker of Pyramix) invented DXD (with Philips). But only the section the audio engineer desires to edit needs to be converted. So, in most cases you still have a mostly “pure” DSD256 file.
It is a small number, agreed, but more than 2-3. The key point is that because both DSD256 (1 bit / 11.3 MEGAhertz) and DXD (24 bit / 352.8 KILOhertz) capture significantly more audible, musical information than all of the other audio formats, many of these recordings are, to my and others’ ears, spectacular.
Cheers,
R
1 Like
If I understand the workflow correctly, anytime there is a level change the data needs to be converted to DXD. I remember looking into this when NaitiveDSD started. Jerad Sacks (sp?) who engineers all channel classics recordings posted a DSD file of the mic capture of a Mahler symphony. It sounded vastly different than the final production. He also challenged listeners to try and hear the difference between 128 and 256
1 Like
Almost all recorded music must be mastered. This applies to both PCM and DSD recordings. My point was simply that both of these formats–DSD256 and DXD–capture significantly more musical information than Redbook CD, and therefore are rightly of interest to many audiophiles. Why would anyone want to start with less of the music? (!)
Of course comparing the total number of Redbook recordings, which have been extant since the early 80s, to a format that has only been around for a few years, makes no sense and will tilt to the former.
The trend in recorded audio is to higher resolution formats and IMHO, all, viable, non-lossy ones (note MQA), should be supported by the top manufacturers.
I hear you for sure! I listen to classical mostly and everything is thing is now high res.
The one thing that bugs me however is that the loudness trend has impacted classical as well. It isn’t horrible but some of the most dynamic classical recordings I have are actually from the late 90s and early 00. Most of these are 16/44 but I bet a bunch of them were actually captured in higher res possibly on a dcs ADC
1 Like
I listen to classical music and cannot think of a single recording where I am aware of the impact of the loudness wars . Please give me some examples so I can try.
Perhaps my main choice of pre-classical repertoire is a factor?
I think the change is quite subtle and tied to mastering engineers adhering to loudness standards developed by streaming services like Apple Music et al. Most of the classical recordings I own from the 90s and early 00s appear to stick with the European R128 standard (below). Many classical recordings today are mastered a bit louder -14 LUFS. I tested a bunch of old and new recordings on my computer with an RME DAC and their software which captures loudness in real time.
In Europe, audio loudness normalization and maximum levels are primarily governed by the EBU R 128 recommendation, aiming for a target loudness of -23 LUFS and a maximum true peak level of -1 dBTP
Thanks . However that has nothing to do with the loudness wars. The latter are ( are) to do with placing the mean ( not peak) recording level as close as feasible to 0dBfs. To do this requires minimal dynamic range. I am not aware of any classical recording subject to this.
NB: I did request some examples so I can try them ( though not measure).
What’s happening however is the target loudness in newer recordings has shifted from -23dB below peak to -14ish dB below peak.
Almost 10dB difference. I would not say every new recording I have is 10dB louder than older stuff but I have noticed that the target loudness of say some newer chamber recordings is louder than ones recorded 20 years ago.
Clearly not everyone follows these guidelines. - some of the newer Chandos recordings are the most dynamic ones I’ve ever heard!
Just a note here…
I have had many conversations regarding file formats - I had a couple at the weekend at the Whittlebury Show where we were asked if it was a “fair” demonstration when we were demoing with high res files when most people don’t own high res files - to which our reply was “Everything you have heard is 44/16 apart from one track which was DSD.” - the assumption was that the clean and dynamic tracks that we were playing were high res and the not-so-clean-and-dynamic ones were 44/16.
The thing is that yes, the higher sample rate and higher bit depth formats are certainly mathematically capable of rendering better resolution but remember that they are simply the container, they describe the bucket that carries your music and nothing more, they are no indication of the quality of the contents.
The bucket can be full of [insert your favourite drink here] or it can be full of vinegar or - more worryingly - it can be full of the cheap generic version of what your favourite drink is (i.e. a lesser format just remapped to fill the bucket).
44/16 is (still) capable of STUNNING quality but the main arbiter of whether something sounds good or not is not (and has never been) the file format (even - IMO - with lossy compressed a la MP3 or a number of other formats although lossy compression does generally have a significant impact) - it is how the music has been post processed and mixed before being laid down in its final form.
I totally get the point that @keiserrg is making and in an otherwise perfect world this would be the case but … 
Phil
1 Like
Good points Phil. Certainly some stunning 44.1 tracks were played at the Varèse demo I attended.
I also have more than one DSD file where I bought the SACD, took the time to rip it, transfer it, and back it up, only to deem the album “unlistenable” (!) : /
Re: file provenance, it’s very important to do your research. Sites like NativeDSD and Eudora Records have precise details about the original recording format.
Sites like HDTracks do not, and you could be listening to MP3 disguised as DXD (!)
A good rule of thumb is that you will never find an original recording in anything higher than DSD256; everything higher has been upsampled.