View Full Version : DSD - The future?
Had a very interesting experience last night courtesy of a friends very high end system listening to music in multiple digital formats on the fly as it were. We are talking ripped CD, 24 bit and DSD. I personally would have liked vinyl in the mix as well as a reference;), however it was not set up as an alternative.
In particular we listened to Diana Krall (The girl in the other room) and Dire Straits ( Love over Gold).
The listening was done via a Chord DSX1000 Reference Network Player and a Chord SPM 1200MK11 Stereo Power amp into KEF Blades. I can tell you this is not a system for the faint hearted, I have heard it many times and it is extremely revealing and can produce monumental dynamics with a magnificent sound stage and a scale only a few of us could hope for? A NAS storage was fed to the Chord Network Player via some Cat7 ethernet cables and all controlled via some nice software via ipad mini.
We used the track Temptation from the Diana Krall album to start with and listened at 16 bit/ 24 bit and DSD version. For me the absolute truth was revealed like night and day between the formats as the KEF Blades are ruthless in their honesty and the DAC in the Chord Player is up there with the best.
First up ripped CD - very good, surprisingly so compared to the other digital formats but switching to 24bit showed its limitations. The 24bit recording had more air and ambience and much greater definition across the recording. Each aspect of the recording could be picked out much more clearly with lots more detail especially around instruments. Very good indeed and difficult to go back to 16 bit. Next up DSD. I was expecting great things considering all the hype around the format but quite simply it did not deliver. I found it slightly dull and lifeless compared to the 24 bit and preferred the 16bit CD rip. DSD did sound quite natural but lost the sparkling detail and ambience of the 24 bit recording.
Dire straits - Private Investigations was the next track we chose to analyse the digital formats and again we had 16bit rip/ 24bit and DSD.
To save a long explanation we found exactly the same differences between the formats with the 24 bit recording sounding glorious. The detail and bite on the acoustic guitar strings was exceptionally well revealed via 24bit, blurred in 16bit and DSD - fail! Just did not measure up. The ambience was also significantly better via 24bit. Again all the instruments had so much more detail and ambient information around them.
One of the conclusions I personally came to other than DSD was a little hyped and unfulfilling, was that DAC,s are not only the key to unlocking the potential of digital audio their development in the future will ultimately bring digital audio to the final frontier - It may one day sound as natural and as real as the best vinyl/analogue systems.:)
Stratmangler
29-08-2015, 11:43
The comparison between PCM and DSD is not necessarily a fair one.
In order to have that particular network player play DSD at all requires a software update.
I have a hunch that the software update allows the player to transcode DSD on the fly to PCM, which in turn means that you may not have been comparing one format against the other in native playback.
Playing back PCM at its native resolution is fairly trivial.
Playing back DSD converted on the fly to PCM requires a lot more processing overhead, which may account for some audible differences.
Firebottle
29-08-2015, 11:47
Interesting James, I have heard some DSD tracks that are mesmerising, mind you that was from a Lampizator Golden Gate DAC.
Can't dispute your findings though with direct comparison, I haven't done that.
As you know I am with you in still preferring a good vinyl setup https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcSUza1FSy_I1vudmyA75kdtATfvKy032 5bV2yVygWp_tpGw5RxO
The comparison between PCM and DSD is not necessarily a fair one.
In order to have that particular network player play DSD at all requires a software update.
I have a hunch that the software update allows the player to transcode DSD on the fly to PCM, which in turn means that you may not have been comparing one format against the other in native playback.
Playing back PCM at its native resolution is fairly trivial.
Playing back DSD converted on the fly to PCM requires a lot more processing overhead, which may account for some audible differences.
Hi Chris, knowing the person who owns this network player I am sure it has had the latest DSD software update.
Regarding my findings listening to the 3 formats I can only report as I find.
I will probably be able to repeat this exercise with a different Network player soon which will be one of the most advanced available. Maybe DSD will perform a little better?
Stratmangler
29-08-2015, 21:53
I'm not defending DSD, just pointing out that the playing field might not be as easily defined as might be thought at first.
The biggy to have to contend with is DSD being just one bit at a time at a frequency of around 2.8MHz, which is a multitude of times faster than the fastest commercially available PCM material at 192KHz .
DSD being just one bit at a time at a frequency of around 2.8MHz, which is a multitude of times faster than the fastest commercially available PCM material at 192KHz .
192000 * 2 * 24 = 9216000 = 9.216Mhz
44100 * 2 * 16 = 1.4Mhz
So DSD is a higher data rate than red book, but less than 96/24
96000 * 2 * 24 = 4.6Mhz
DSD uses noise shaping to get the data rate down from what would be required for simple delta sigma encoding of 96/24 while retaining much of the performance.
Stratmangler
30-08-2015, 09:40
192000 * 2 * 24 = 9216000 = 9.216Mhz
44100 * 2 * 16 = 1.4Mhz
So DSD is a higher data rate than red book, but less than 96/24
96000 * 2 * 24 = 4.6Mhz
DSD uses noise shaping to get the data rate down from what would be required for simple delta sigma encoding of 96/24 while retaining much of the performance.
I meant that just the clocking speed is waaayyyy faster with DSD.
The data rate equates to 24/88.2, going off the options for playback using Foobar.
I meant that just the clocking speed is waaayyyy faster with DSD.
The data rate equates to 24/88.2, going off the options for playback using Foobar.
Yes, but as I was pointing out, thats only true if you are talking about the word clock. The bit clock for any serial PCM stream (i2s, S/pdiff toslink) is higher as the data is sent a bit at a time as well.
24 * 88200 * 2 = 4.2Mhz, foobar is probably translating DSD into that rate of PCM.
But it depends what you mean by data rate I guess.
Stratmangler
30-08-2015, 10:27
Yes, but as I was pointing out, thats only true if you are talking about the word clock. The bit clock for any serial PCM stream (i2s, S/pdiff toslink) is higher as the data is sent a bit at a time as well.
24 * 88200 * 2 = 4.2Mhz, foobar is probably translating DSD into that rate of PCM.
But it depends what you mean by data rate I guess.
I was just taking an overly simplistic view and referring to the word clock - that'll teach me :doh: :D
The figures shown in Foobar are DSD to PCM conversions.
I'd suggest that if DSD sounded so poor then there was something badly wrong somewhere.
DACS are advancing and getting better and better? It's an oft heard cry but a listen to my 1989 Technics SLP1200 tells me that isn't the case. Forget the whizz-bang pyrotechnics of hi-rez number crunching - it is the implementation that matters most.
I'd suggest that if DSD sounded so poor then there was something badly wrong somewhere.
DACS are advancing and getting better and better? It's an oft heard cry but a listen to my 1989 Technics SLP1200 tells me that isn't the case. Forget the whizz-bang pyrotechnics of hi-rez number crunching - it is the implementation that matters most.
The DSD recordings were certainly not as good as 24 bit. There was nothing badly wrong with the system as it is set up way beyond most of us would be able to achieve:eyebrows:
Simply DSD is not all its cracked up to be. I have heard it before and not been impressed even though one or two recordings were good.
I can assure you the system I was listening to would resolve exactly what is going on. Maybe the mastering of the DSD was not well executed on the recordings I chose to compare?
Anything including or bettering 44.1KHz 16 bit can sound great.
I tend to notice DSD somehow sounding smoother than multibit as a rule. Not always a good thing though.
Disturbingly some rock can sound better compressed than at higher bit rates I find. But then rock relies on distortion so no surprise.
Anything including or bettering 44.1KHz 16 bit can sound great.
I tend to notice DSD somehow sounding smoother than multibit as a rule. Not always a good thing though.
Disturbingly some rock can sound better compressed than at higher bit rates I find. But then rock relies on distortion so no surprise.
Yes DSD does sound a bit smoother but on the recordings I heard it lacked bite and detail, in fact it sounded quite boring compared to 24bit. Even the CD rips were good but could not match 24 bit for the air and greater ambience around instruments. The 24bit recordings also had more presence and certainly bought you closer to the recording.
With the track Private Investigations you could actually hear a plectrum on the strings of the guitar easily where as on the 16 bit it blurred into the mix.
I think classical people like it because it smooths out any roughness on strings.
Smoothness means lack of bite as you say.
DSD is not the future it is only gaining ground as a computer and a laptop is a more popular source for many.
Martin Colloms never seemed keen on it TBH and I'm pretty sure he also thought it lacked pace/rhythm years ago. That is probably just due to a less incisive sound, though, I'd wager.
I think classical people like it because it smooths out any roughness on strings.
Smoothness means lack of bite as you say.
DSD is not the future it is only gaining ground as a computer and a laptop is a more popular source for many.
Martin Colloms never seemed kean on it TBH and I'm pretty sure he also thought it lacked pace/rhythm years ago. That is probably just due to a less incisive sound, though, I'd wager.
From what I have heard 24 bit is very good indeed and with every advancement in DAC technology it will one day may sound as good as vinyl:) I view this as similar to digital photography where the pixel rate went up but more importantly the processing got better. Look what apple are now achieving with their iphone camera with a few pixels and smart processing.
Although I am not very knowledgeable about many of the aspects of digital recording, I have read plenty. One area thats sticks out for me is the time element in digital processing ( clocking).
If you are reading this Nick G maybe you could enlighten me more but I remember talking with you and you mentioned the research on how we hear in patterns and not just waveforms and this maybe the key to aligning digital processing to smooth out the lags or minute time delays in processing the bits?
Vinyl is liked, I think, because of noise and distortion injection that many people like.
That's why I use a technically questionable DAC. Its triode output stage adds something I like to hear. What's more, the nature of that "addition" is easily variable by valve rolling.
In all honesty it beats my vinyl setup more than my vinyl setup beats it. High bit rate/depth or not.
cyclopse
30-08-2015, 20:29
Vinyl is liked, I think, because of noise and distortion injection that many people like.
That's why I use a technically questionable DAC. Its triode output stage adds something I like to hear. What's more, the nature of that "addition" is easily variable by valve rolling.
In all honesty it beats my vinyl setup more than my vinyl setup beats it. High bit rate/depth or not.
When vinyl is done right it's a tough act to beat. Bit like getting a bit more of everything.
There's a fly in the ointment here -- the recordings mentioned surely aren't DSD masters! They both originate from PCM recordings -- an early 80s one in the case of Love Over Gold (my Vertigo LP has that 'digital recordimg' logo on the front) which will be 16-bit originally.
To do this fairly to DSD you need native DSD recordings, which can then be converted to PCM for comparison. Difficult to know how much impact this would have on the PCM in this case.
There are differences between devices in conversion from DSD to PCM. When I bought my Pioneer blu-ray player it was set to decode DSD to PCM. Switching to DSD output improved the sound to a surprising degree. I'm sure my Yamaha is still converting to PCM (it has to for bass management in my sub-less system), and it clearly has much more computing "grunt" than the Pioneer as it shows it a clean pair of heels.
To do this fairly to DSD you need native DSD recordings, which can then be converted to PCM for comparison.
Not sure if thats still a fair comparison, better the DSD sent via DSD on USB to a native DSD DAC, converting to PCM anywhere in the chain is missing the point IMHO.
Not sure if thats still a fair comparison, better the DSD sent via DSD on USB to a native DSD DAC, converting to PCM anywhere in the chain is missing the point IMHO.
You have to convert to PCM to compare, or identify the effect of PCM, surely? To compare the two systems with the same recording.
Unless you made a test analogue recording of a live event, mixed it and mastered it separately (but identically, if you get my meaning) to PCM and DSD. Are there any albums out there from analogue masters that have been given separate DSD and PCM releases from the same master that could be used for such a test?
Maybe I'm being too scientific and it's enough to simply express a preference, Whiskas style...
Not sure if thats still a fair comparison, better the DSD sent via DSD on USB to a native DSD DAC, converting to PCM anywhere in the chain is missing the point IMHO.
I had meant to say played as pure DSD and then played with conversion to PCM to compare...
When I did this I used Steely Dan - Gaucho - the 2001 remaster on CD, the DVD-A 24/192 and the SACD. Using an Oppo player. I'm not sure it is the same master but they all sounded very similar. That was not on my system but on a friend's. The DSD was not noticeably poorer in any respect. Were they all identical? No, I don't think so, but I did think I would struggle to tell them apart blind without some practice.
It is very difficult to know what you are really listening to .e. what processes any source has really been through. I just recommend reading Wiki on DSD (https://en.wikipedia.org/wiki/Direct_Stream_Digital) and forming some opinions in your own mind.
Thats the problem with any such a/b comparison, you have no way of knowing if they are originally the same.
You have to convert to PCM to compare, or identify the effect of PCM, surely? To compare the two systems with the same recording.
But you are still then not sure if you are hearing a difference that is due to the difference between DSD and PCM or the difference is caused by the conversion process. The only way would be to make your own recordings using something like this http://tascam.com/product/da-3000/
You have to convert to PCM to compare, or identify the effect of PCM, surely? To compare the two systems with the same recording.
Unless you made a test analogue recording of a live event, mixed it and mastered it separately (but identically, if you get my meaning) to PCM and DSD. Are there any albums out there from analogue masters that have been given separate DSD and PCM releases from the same master that could be used for such a test?
Maybe I'm being too scientific and it's enough to simply express a preference, Whiskas style...
I know where your coming from Malcolm but not sure how much DSD material is out there to do these comparisons. However I will listen to some native DSD recordings and assess.:)
How can you tell you are listening to a true DSD recording that has not been converted to PCM or mangled in some other digital configuration.:scratch:
[QUOTE=lurcher;680240]Thats the problem with any such a/b comparison, you have no way of knowing if they are originally the same.
This has always been my suspicion regarding a lot of the digital material floating around out there and it has been proven that what people think they have is sometimes not the version they thought the had.
This may account for the reason why I did not find the DSD recordings particularly good.
How can you tell you are listening to a true DSD recording that has not been converted to PCM or mangled in some other digital configuration.:scratch:
You can't unless you can find detailed information on the provenance of the recording/mastering and that is not so easy. Or make you own recordings as suggested earlier.
Chord Dacs do no playback DSD natively, they convert to high rate pcm and most decimate. The new upcoming Dave is touted not to decimate DSD, but still will likely have a lot of dsp processing, which is not very good for the delicate dsd signal.
Essentially, the OP was listening to good pcm and bad PCM and though that was DSD.He should try to hear DSD on evan a basic Lampi and he will finally get it. Also, ensure the recording was an analog transfer or native DSD recording, as that will beat a regular pcm file converted to dsd.
Also, do a search at WB Forum to find at least 4 album simultaneously recorded in DSD and PCM at the same time.
For the record, there are tons of DSD out there.Once you hear DSD done right, you will GET it. The UK has been far too resistant for too lang and is missing out on a whole load of goodness.
Justin, take a trip to Greg and let him play you some of the Opus Sampler albums! You have a stunning DSD engine in your Dac and I am shocked you dont use it. I know 2 guys with Trinity Dac who also bought Lampi Golden Gates for DSD playback.
Audio Advent
02-09-2015, 01:48
192000 * 2 * 24 = 9216000 = 9.216Mhz
44100 * 2 * 16 = 1.4Mhz
So DSD is a higher data rate than red book, but less than 96/24
96000 * 2 * 24 = 4.6Mhz
DSD uses noise shaping to get the data rate down from what would be required for simple delta sigma encoding of 96/24 while retaining much of the performance.
You are out by a factor of 2 here: the DSD data rate is for a single channel. DSD64 is exactly the equivalent data rate of a single channel of 16 bit/176.4kHz = 2.8224Mhz.
Then there is DSD128 and DSD256, doubling the data rate each time.
I suggest you to read this article and especially short, but honest Mark Levinson's comment at the bottom of the page: http://www.stereophile.com/content/jared-sacks-dsd-present-and-future-page-2#5bS4BMeyGBFcqt4r.97
You are out by a factor of 2 here: the DSD data rate is for a single channel. DSD64 is exactly the equivalent data rate of a single channel of 16 bit/176.4kHz = 2.8224Mhz.
Then there is DSD128 and DSD256, doubling the data rate each time.
Ok, I agree, I had forgotten that there was a left and right data stream with DSD, But 1.4 < 2.8224 < 4.6, so its still valid. My point was that once PCM becomes serial its data rate is higher than was suggested. I should have remembered that given the time I spent getting DSD switching to work correctly with the DAC I make.
Lurcher, but isn't dsd fundamentally different to PCM, which is word sample based, where DSD is simple change to previous refence point?
PCM sample defines what the actual voltage is each time, while DSD simply says, go up or won from the previous state. DSD is thus much more data efficient, ie less redundant info. This makes simple data rate evaluation an apples to oranges comparison.
I suggest you to read this article and especially short, but honest Mark Levinson's comment at the bottom of the page: http://www.stereophile.com/content/jared-sacks-dsd-present-and-future-page-2#5bS4BMeyGBFcqt4r.97
Interesting article Krzysztof.
I like Mr Levinsons last point 3. Record in DSD, convert to analog, master in analog, and convert to DSD.
I love this bit:
When you listen to PCM, you can literally hear it as a block of sound coming out of the speaker.
Yes, sure you can.
if you are going to covert it to analogue to muck about with it you might as well just record in analogue in the first place. Either way you reduce bandwith and dynamics and add noise.
And this bit:
Serinus: Who is recording in native DSD?
Sacks: Pentatone, Harmonia Mundi, BSO, and Alia Vox for starters
So pretty much nobody then. Everyone else is recording in PCM then occasionally converting to DSD so they can release it on DSD for the niche market that think they care about such things. So, what exactly is the point?
Lurcher, but isn't dsd fundamentally different to PCM, which is word sample based, where DSD is simple change to previous refence point?
PCM sample defines what the actual voltage is each time, while DSD simply says, go up or won from the previous state. DSD is thus much more data efficient, ie less redundant info. This makes simple data rate evaluation an apples to oranges comparison.
Yes, but I was originally replying to a simple question about clock rate. You could argue (as you indirectly do) that DSD is actually a lossy compression method that produces a delta sigma bit stream that can be reconstructed to an equivalent sounding result.
You could transfer any binary file as PCM and reconstruct the original unchanged (assuming no data transfer problems), you cant do the same with DSD.
DSD is only more efficient because of the compression (noise shaping), not the fact that its one bit.
I love this bit:
When you listen to PCM, you can literally hear it as a block of sound coming out of the speaker.
Yes, sure you can.
if you are going to covert it to analogue to muck about with it you might as well just record in analogue in the first place. Either way you reduce bandwith and dynamics and add noise.
And this bit:
Serinus: Who is recording in native DSD?
Sacks: Pentatone, Harmonia Mundi, BSO, and Alia Vox for starters
So pretty much nobody then. Everyone else is recording in PCM then occasionally converting to DSD so they can release it on DSD for the niche market that think they care about such things. So, what exactly is the point?
Exactly my thoughts Martin!:)
.Justin, take a trip to Greg and let him play you some of the Opus Sampler albums! You have a stunning DSD engine in your Dac and I am shocked you dont use it. I know 2 guys with Trinity Dac who also bought Lampi Golden Gates for DSD playback.
I didn't order the DSD module as the original machine I was supplied for demo purposes with did not work in DSD mode. Converting PCM to DSD on the fly in JRiver seems counter-intuitive to me. I do not listen to classical music - which is where most DSD seems to be at. Also, I owned an SACD player and I wasn't that impressed with SACDs.
I know, I know, I am wrong, LOL, but I am not going to order the module, Norman. Very happy with PCM, though by all accounts those who have working DSD modules seem impressed I don't disagree.
Interesting article Krzysztof.
I like Mr Levinsons last point 3. Record in DSD, convert to analog, master in analog, and convert to DSD.
Yes, it's interesting indeed. For me DSD is technology to keep an eye of course, but well done PCM not lacking anything technically and could be used in the whole recording/mixing/mastering/publishing chain which is impossible with DSD by its nature being 1-bit.
We need to keep in mind as well - kinda OT, that high resolution should be legit if the whole mentioned chain is done in at least 24/88.2kHz. Big chunk of the files offered by Hi Rez friendly websites are transferred from the source which was NOT able to capture in high resolution. Another point to buy this high resolution contenders is they recently usually are done in different mastering, usually better than their CD version from the past early CD era - not always, but before buy take some time to research who mastered it, when and what was the provenance of the file.
So, as the summary, one could say, that we don't need yet another format, we just need to take of the recording/mastering processes mitigating any traces of the loudness war - just dreaming….
Serinus: Who is recording in native DSD?
Sacks: Pentatone, Harmonia Mundi, BSO, and Alia Vox for starters
So pretty much nobody then.
Guess you missed the 'for starters' bit ... easy mistake to make when you read selectively :lol:
anubisgrau
12-12-2015, 23:45
just returned from a lampizator DSD demo. it looks like a real game changer to me. not sure if this kind of sound was possible before from digital within that price range (RRP 1800e).
Audio Advent
13-12-2015, 01:52
Very nice.
I was learning about Meridian's MQA lossless, what it does and the sonic impressions so far. In a nutshell, apart from the data compression, it compensates for pre and post ringing in the filters from the A/D conversion process by keeping a database of converters and compensating appropriately, much like a lens compensation in digital cameras. The reported sonic result is one of a very very clean and precise sound, sounding so much more like real life.
DSD doesn't suffer from pre or post ringing in the A/D process by it's nature... and so people's description of DSD's sound is one of a clean, open and precise, much more as real life.
Seems like DSD is the simpler way of getting that same clean and open sound.
Agree that DSD often sounds more "natural" (analog?) than its PCM brethren.
But the Meridian MQA advent seems to me like bollox... Of course we have better equipment to store music digitally today and digitizing analog tapes with better equipment are all good. But, restoration of early A/D conversions cannot, ever, recreate information that was there in the first place but lost when first A/D converted.
As for the replay part of the concept, i personally havent heard the result, but well selected recordings can always be used to convince.
Yes, i realise that some form of intelligent compression is being used to squeeze 24/96 worth of info, into a 24/44.1 format. That does NOT mean the result is more highly resolved than 24/96 which is probably more than most need/can distinguish.
Nah, DSD seems a more natural way to go. If Sony hadn't been such moreons and copy protected it like they did it would have been a standard today!
If pre and post ringing was actually an issue in reality then most of your vinyl records will suffer from it as they have all been digitised before they were cut. Since no-one complains about this or is campaigning for a return to analogue cutters then we can assume it is a non-issue, or at best a problem for the DAC designer to sort out.
Regardless of whether the recording is on vinyl, CD, DSD, 24/96 or whatever it is still simply a record of voltage change over time - wavelength and amplitude, that is all that is on there. Those instructions are read and the signal passed to your amp which instructs your speaker drivers to move in and out - that creates sound pressure waves which reach your ears and then - somewhere in your brain - those signals are pattern-recognised and converted to what we call 'music'.
The way I see it if you want better sound you are better off spending the money on better kit - especially the speakers, rather than re-purchasing your record collection in some 'new' format.
As for the replay part of the concept, i personally havent heard the result, but well selected recordings can always be used to convince.
Spot on and I think this needs to be recognised - these people are after your money and they won't hesitate to skew things in their favour if they think it will bring in sales. Just because this is hi-fi it doesn't mean no-one is trying to make money out of it.
Rothchild
13-12-2015, 10:11
I was learning about Meridian's MQA lossless.... it compensates for pre and post ringing in the filters from the A/D conversion process by keeping a database of converters and compensating appropriately,
Gotta reference for this? It was mentioned last time DSD came up and I could find no evidence for it being the case, AFICT (and in broad strokes) MQA is something like 24/96 but with the (minimal) spectral content above 20kHz subject to some lossy compression.
Stratmangler
13-12-2015, 10:33
If Sony hadn't been such moreons and copy protected it like they did it would have been a standard today!
How can Sony have been morons and copy protected it?
They didn't copy protect it.
They just made sure that SACD capable drives didn't make it out into the computer world at large.
The one loophole they left was via the PS3, and lo and behold, someone managed to exploit it, hence all the PS3 ISO rips that can be found on the 'net.
Sony's mistake was not pushing the fact that music encoded on the format was secure because nobody had a computer drive that could read them, and they didn't tout it around the record labels as such.
If they'd done that then it might have been a success for them.
And none of us would have SACD capable PC drives, so all the music released on the format would still be securely locked to the discs that store the information.
It's all well and good coming out with platitudes after the event, but not one of us has a crystal ball to predict the future ;)
Audio Advent
13-12-2015, 13:26
How can Sony have been morons and copy protected it?
They didn't copy protect it.
They just made sure that SACD capable drives didn't make it out into the computer world at large.
The one loophole they left was via the PS3, and lo and behold, someone managed to exploit it, hence all the PS3 ISO rips that can be found on the 'net.
Sony's mistake was not pushing the fact that music encoded on the format was secure because nobody had a computer drive that could read them, and they didn't tout it around the record labels as such.
If they'd done that then it might have been a success for them.
And none of us would have SACD capable PC drives, so all the music released on the format would still be securely locked to the discs that store the information.
It's all well and good coming out with platitudes after the event, but not one of us has a crystal ball to predict the future ;)
They certainly prevented digital outputs giving out DSD on SACD machines by way of licencing agreement. You can have DSD inputs and outputs as it's just a standard but SACD itself is licenced so there are rules to obey.
But SACD is dead because we can send and download the files without the shiny disc element or the licencing.
Audio Advent
13-12-2015, 13:43
Gotta reference for this? It was mentioned last time DSD came up and I could find no evidence for it being the case, AFICT (and in broad strokes) MQA is something like 24/96 but with the (minimal) spectral content above 20kHz subject to some lossy compression.
There is this
http://youtu.be/T5o6XHVK2HA
which has been verified by Bob Stuart for technical accuracy.
The analogy used is one of alogrithms correcting for lens errors in digital cameras, like chromatic aberration. A database is kept in the camera of all the errors created by the specific lens in the camera and so algorithms are applied to the image to correct them. The video above says that the MQA encoding process (not something on the consumer side) is able to work out the signature of the A/D converters used from noise distribution and then apply the appropriate time-smearing correction for that converter.
Meridian have been working for at least 3 years with studios developing MQA so I imagine a lot of that will be sorting out and logging the correction for common converters. It might be that an obscure studio using an obscure converter won't get anywhere with MQA sonically but then again maybe the studios need to be licenced in the first place so small ones won't get a look in.
And no lossy compression - it's more like Meridians 1990's dither algorithms which manage to put data into the noise floor area where the data is otherwise wasted.
Audio Advent
13-12-2015, 14:01
If pre and post ringing was actually an issue in reality then most of your vinyl records will suffer from it as they have all been digitised before they were cut. Since no-one complains about this or is campaigning for a return to analogue cutters then we can assume it is a non-issue, or at best a problem for the DAC designer to sort out.
I certainly never complain about FM but I know it doesn't sound as good as CD after having heard CD. I never complained about cassette. People don't complain about their Bose systems.. Some people, after hearing the Hypex Ncore 400 Class-D amps can't go back to their previous Class-A monsters because they don't sound as good. Until you hear something new then why would you complain about your current set-up?
Vinyl is going to have a LOT more ringing due to all the filtering going on both in the RIAA stage and the bandwidth filtering before the cut, not to mention the digital masters. All filters of any kind cause ringing, even speaker crossovers. This only tries to eliminate an extra set of ringing specific to PCM digital and apparently it works according to those who have heard it and suitably can recognise the effects of ringing.
If everything, like you say, suffers from pre and post ringing ... then why would anyone know any different to be even able to complain? I bet though the results are subtle even if significant.
It's not an issue for DAC designers though, it is an issue for ADC designers (it's about ringing in the A/D anti-alias filters mostly). That is why some A/D converter designers in the pro world have advocated hi-res recording of as much as 384KHz or at least DXD, and some DSD - for exactly that issue for the last 10 years or more. Yet I seem to remember in a high-res discussion that their very same reasons for pushing high-res were pooh poohed as just marketing BS. Now you're saying that if it was a problem they would be doing something about it... well, err.
I'm not advocating MQA myself - I doubt I will ever bother to get into it if it's not handed on a plate - but your argument of why ringing isn't a possible sonic problem doesn't stand up.
Audio Advent
13-12-2015, 14:13
Spot on and I think this needs to be recognised - these people are after your money and they won't hesitate to skew things in their favour if they think it will bring in sales. Just because this is hi-fi it doesn't mean no-one is trying to make money out of it.
So how come all your hifi isn't from about 1910? All that came after the wax cylinder is just marketing you've fallen for surely?
Some people SAY that you can get a better sound with 78 vinyl than a wax cylinder but it's just a ploy to get me to buy my whole collection again on a new format!!
and so it goes with each sonic improvement either via new design or new technology. Where is this mythical line where suddenly the improvements were no longer honerable and suddenly marketing tricks and BS? Is it a personal line where your attitude changed or is there a technological time line where you can say before there were genuine improvements and after it was all smoke and mirrors? Or is it a point nothing to do with sound but just were there marketing changed to a style you don't like or trust?
I can't get my head around why some things in your mind might be sonic improvements (when you upgrade your hifi) and why at other times, improvements are simply not possible over 70s technology...
StanleyB
13-12-2015, 15:45
So how come all your hifi isn't from about 1910? All that came after the wax cylinder is just marketing you've fallen for surely?
First you have to make clear what you consider to be hifi. Hifi reproduction didn't hit the market, and the ears or audio enthusiasts, till the 40's. So that would put you out by three decades.
I can't get my head around why some things in your mind might be sonic improvements (when you upgrade your hifi) and why at other times, improvements are simply not possible over 70s technology...
Some things are pretty much as good as they are going to get without a big technological leap forward. Which would probably occur in a domain other than hi-fi and be applied to hi-fi at a later date.
That's why we see so much pecking around the edges with 'hi res' and so on. I simply don't see the logic in trying to push an extra 0.1% out of digital when most loudspeakers won't deliver it in any case. Really good loudspeakers are very expensive because they require mechanical engineering so assuming everything else is already up to snuff that is really the area to explore if you want an improvement in sound you don't have to strain to -possibly- hear. New formats that maybe eliminate ringing we can't detect and maybe have sound at such high frequencies we cannot hear it don't really seem to be worth expending cash on if you don't already have speakers that are the absolute best available. It's fiddling whilst Rome burns.
StanleyB
13-12-2015, 19:37
Some things are pretty much as good as they are going to get without a big technological leap forward. Which would probably occur in a domain other than hi-fi and be applied to hi-fi at a later date.
The question is, do we really want to hear the sound of a bank of Marshall speakers in our front room? The quality of the equipment many of us have forked out on surpasses the quality of the sound of the equipment used on many of the tracks we so enjoy listening to. I am not interested in the kind of accuracy that would take me backwards in sound quality.
I was thinking more of technological developments in terms of space exploration, or military, the sort of thing that has big research budgets. Then that trickles down into everyday uses, one of which might be hi-fi.
I don't want playback through Marshall stacks either :)
StanleyB
13-12-2015, 20:10
I don't want playback through Marshall stacks either :)
Unfortunately too many people are overlooking this possibility as they invest heavily into the promise of "real life" experience "as if you were there at the recording".
It crossed my mind ages ago that in order to reproduce the life like sound of a guitar, you would have to design a piece of equipment that could reverse engineer the distortions introduced by the various mics, recording equipment, and the playback equipment itself, before you could get within miles of that sort of accuracy. But it just so happens that Texas Instruments brought out a DAC chip PCM5142, and a piece of software called PurePath™ Studio that comes very close to that process.
Rothchild
13-12-2015, 20:40
Unfortunately too many people are overlooking this possibility as they invest heavily into the promise of "real life" experience "as if you were there at the recording".
It crossed my mind ages ago that in order to reproduce the life like sound of a guitar, you would have to design a piece of equipment that could reverse engineer the distortions introduced by the various mics, recording equipment, and the playback equipment itself, before you could get within miles of that sort of accuracy. But it just so happens that Texas Instruments brought out a DAC chip PCM5142, and a piece of software called PurePath™ Studio that comes very close to that process.
I can't believe it's already 15 years since Antares (most famous for being first to market with a workable Autotune product) released their mic modeller: http://www.soundonsound.com/sos/apr00/articles/antares.htm so part of your vision is doable, but realistically (unless we're only talking about single instrument recording) you're right because the bottom line is once it's mixed the cake baked and we can't unbake it.
Even with the solo guitar example, the first part of the recording chain (and the distortions therein) is the instrument and the room it's recorded in I'm sceptical that just applying retrospective phase/eq changes (to account for the rest of the chain) could be done so you just had the 'instrument' (guitar and room)
I'd also like to think (as a recordist) is that if you did hear the basic version, you'd appreciate the choices that go on to choose a mic and signal path to really make it 'sound like a guitar'
Stratmangler
13-12-2015, 20:49
The question is, do we really want to hear the sound of a bank of Marshall speakers in our front room? The quality of the equipment many of us have forked out on surpasses the quality of the sound of the equipment used on many of the tracks we so enjoy listening to. I am not interested in the kind of accuracy that would take me backwards in sound quality.
I wouldn't want to have a Marshall amplifier turned up to serious levels in my front room, but I'd like to have an accurate rendition of it at sensible volume levels.
A couple of years back I went to The Band On The Wall in Manchester to see the Antonio Forcione Trio performing.
The sound in the venue was sublime, and the performance and virtuosity of the musicians was jaw dropping and awe inspiring.
The thing that struck me the most about that gig was that the sound was not very far away from the way it sounds playing that CD at home - it sounded like the band were playing through a decent hifi system. And that included the one tune where they turned off the sound reinforcement equipment because of the pickup on one of the instruments was broken - that's right, the sound was purely that of the acoustic instruments on their own, and there was very little (ie next to no) difference in volume level from the stage.
Throughout the gig they had used the bare minimum of sound reinforcement.
Hifi can do realistic playback, but the circumstances when it can are few and far between, unless of course your listening tends to be small acoustic combos and ensembles.
Audio Advent
13-12-2015, 23:51
I want a Star Trek holodeck basically, aurally speaking. And when you hear a Marshall stack on stage perhaps during soundcheck, what makes it real is not the Marshall stack (which is what the mic picks up) , it's the stereo echo around the venue or the delay from the delay line from far speakers and the sound of the wind and stuff if an outside gig with perhaps some clattering as people set up some steel deck or scaff poles.
The problem is that someone thinks that the good thing about sound is only about the notes .. so they've stuck the bloody mic in front and tried to isolate it from everything else - a couple of ambient mics would be MUCH more like it.
But that's my taste and hence I'm looking for realism - just jolly difficult to find recordings which set out to achieve the same thing. In the meantime I try to at least get the hifi together which will recreate what I want. DSD or very high res will be more likely to achieve that in my opinion because of impulse response and the lack of time smearing, comparatively speaking.
Audio Advent
13-12-2015, 23:57
So that's what I want my hifi to be CAPABLE of ... but I accept and enjoy greatly music in all it's forms including the constructed non-live events, the sonic events or sonic constructions which are most of the albums I listen to, especially electronica for which my ideal hifi's capabilities will not be fully used.
Audio Advent
14-12-2015, 00:02
But it just so happens that Texas Instruments brought out a DAC chip PCM5142, and a piece of software called PurePath™ Studio that comes very close to that process.
Is that what is used on miniDSP digital crossover boards and the like? Just connecting "miniDSP" as mentioned in a PurePath youtube video with the name of the digital crossover product.
So that's what I want my hifi to be CAPABLE of ... but I accept and enjoy greatly music in all it's forms including the constructed non-live events, the sonic events or sonic constructions which are most of the albums I listen to, especially electronica for which my ideal hifi's capabilities will not be fully used.
You should have been at the last NEBO. Gordon's (Halfway Tree) system was as close to that 'holodeck' as I have ever heard.
robtweed
14-12-2015, 16:02
I want a Star Trek holodeck basically, aurally speaking. And when you hear a Marshall stack on stage perhaps during soundcheck, what makes it real is not the Marshall stack (which is what the mic picks up) , it's the stereo echo around the venue or the delay from the delay line from far speakers and the sound of the wind and stuff if an outside gig with perhaps some clattering as people set up some steel deck or scaff poles.
The problem is that someone thinks that the good thing about sound is only about the notes .. so they've stuck the bloody mic in front and tried to isolate it from everything else - a couple of ambient mics would be MUCH more like it.
But that's my taste and hence I'm looking for realism - just jolly difficult to find recordings which set out to achieve the same thing.
So one of my hobbies is live audio recording, where you're trying to capture not some clinical studio multi-miked reconstruction of the musicians and their instruments, but more closely what it sounded like to be there, complete with ambience - kind of warts and all stuff. The key is simple mic technique and as little processing as possible. You learn a lot about the other end of hi-fi audio doing this - ie the devices you need to get a high-res capture of the original sound. These days that's really just 2 things: the microphones and the recording device - known as field recorders.
The mics people use for this kind of thing tend to be high-end ones from the likes of Schoeps, Neumann or DPA, but there are some real bargains these days from small outfits such as Line Audio in Sweden - £99 for a mic that will seriously rival the German and Danish old-guard.
As to field recorders, the one that is setting this world alight is the new Zoom F8, with clean, low-noise mic pre-amps and A/D converters to rival the previous king of the heap Sound Devices, only for £800 instead of £5000 or more for SD gear.
Here's an example - a track from an amateur big band's Christmas concert I recorded last night at a small village hall:
https://s3-eu-west-1.amazonaws.com/robs-music-files/wishumerryxmas.wav
What you hear is from 2 DPA 2011C cardioid mics and 2 Line Audio OM1 omni mics arranged as a phased array. Google "Tony Faulkner phased array" and you'll find the details of what this arrangement aims to achieve, but suffice to say the omnis beef up the sound and add in the ambience and space that the ORTF-arranged cardiods lack. It was recorded directly to my Zoom F8. All I did was mix down a balance between the two mic pairs from the F8 using Audacity - otherwise no processing.
Back to the OP - getting realism is as much how it's recorded in the first place as the playback gear and resolution beyond CD quality. Originally recorded at 192k/24, I mixed the sample here down to 44.1/16 simply to reduce the download size - it actually doesn't lose a lot. However, if anyone wants a high-res copy to compare, ping me.
Of course it's a very different sound from a studio-recorded close-miked, remixed and reconstructed big band, with all the mistakes edited out and replaced, but I think it's a lot closer to what a big band actually sounds like when you're there in front of it.
If nothing else, hopefully I provided a bit of Christmas spirit with this track!
Interesting. I might invest on one of these recorders. A friend was playing me back a recording of his band he had made using a Zoom - but he had just placed it on an amp at the back of the stage behind the drum kit. Vocals came out okay, as did the drums, but the bass guitar didn't come through too well. It was still quite impressive for what it was, though.
robtweed
14-12-2015, 17:20
That was probably something like a Zoom H4n - a pretty nice hand-held field recorder (I have one but probably going to sell it now).
The F8 is in another league. Previous Zooms were heavily criticised for poor, noisy mic pre-amps and flimsy design. I think they realised there was a huge gulf in the marketplace between the basic but effective hand-held field recorders (a few hundred pounds) and the high-end pro market which Sound Devices (£3k - 6k or more) had all to themselves for years, so Zoom upped their game and aimed for as close to Sound Devices quality as possible, but with a psychologically significant $999 price tag. As a result, F8s are extremely popular and pretty hard to come by right now as a result, but they're an awesome and amazingly versatile piece of kit, so worth getting your hands on if you can find one.
Rob is it possible to use the H4N to record output from the desk? What connection would be needed?
robtweed
14-12-2015, 18:28
No, Martin
The H4n has some limitations and this is one: it doesn't have line level inputs. It can be done with line level attenuators:
https://www.youtube.com/watch?v=6rQIt9t96p0
but that's a bit of a pain (and I'm not sure what those attenuators cost).
The F8 is much more versatile. All 8 inputs are combined mic/line inputs - a 1/4 inch jack is automatically detected as a line level source. I used that last night to take 2 aux out feeds from the mixing desk used for the vocalists' mics for the PA, so I could beef up their levels at times when they sang - a really nice feature of the F8.
However, this is a very nice alternative field recorder to the H4n (though considerably more expensive):
http://www.thomann.de/gb/sony_pcm_d100.htm
If you're happy to use its built-in mics and otherwise just need line and/or optical digital inputs, then for about £450 it's a great way to record very high quality, hi res (and even DSD) audio. If the F8 hadn't come out I'd probably have bought one.
Rob
Thanks Rob. That doesn't look like too much trouble with the attenuators. I'll look into it.
robtweed
14-12-2015, 18:39
By the way, if you want to do DSD recording, this is probably one of the coolest devices around right now:
http://tascam.com/product/da-3000/
https://www.musicmatter.co.uk/tascam-da3000?gclid=CKrW8vD-28kCFcE_Gwod5loO5w
It's not really portable - it's a rack-mounted device more suited to the static environment of a studio, but people are using these for all kinds of purposes - including vinyl LP archiving to DSD
Rob
Audio Advent
15-12-2015, 18:29
By the way, if you want to do DSD recording, this is probably one of the coolest devices around right now:
http://tascam.com/product/da-3000/
https://www.musicmatter.co.uk/tascam-da3000?gclid=CKrW8vD-28kCFcE_Gwod5loO5w
It's not really portable - it's a rack-mounted device more suited to the static environment of a studio, but people are using these for all kinds of purposes - including vinyl LP archiving to DSD
Rob
The Korg MR1 is probably the cheapest option for DSD and well regarded too. Secondhand they can be had for £75 - £130 odd, when about £800 originally. Discontinued a few years ago now. Great meters on screen and easy to set up, even has balanced capability with an adapter and good mic-pres. No built in mic or anything, pure recorder/playback. And internal HD only... I guess these restrictions make it perfect for some but also why they're so cheap now despite their recording quality.
There was a corresponding MR-1000 rack version with DSD256 and better I/O. Then replaced by the MR-2 and MR-2000 with SD options and higher rates (which might now also be discontinued).
Audio Advent
15-12-2015, 18:33
You should have been at the last NEBO. Gordon's (Halfway Tree) system was as close to that 'holodeck' as I have ever heard.
I'll have a look for some posts about it...
EDIT: well I just did! And miniDSP + DIY active speakers is exactly where I'm headed in the long term too for perhaps all the same reasons! I'll have to check out the Linkwitz LX521 designs a little more.
Makes me think that sorting out the speakers and their control via DSP crossovers can resolve so many of the problems we try to solve with sources and amps.
Of course DSD is incompatable with any form of DSP - it has to be some form of PCM to do the processing.
Ali Tait
15-12-2015, 18:54
I'm certainly happy with mine on my OB's.
Interesting reading this post a few years on now the digital formats seemed to have settled down with 24 bit hirez now accepted as the standard for high resolution digital audio and DSD seems to have just faded away and died, probably due to lack of available material and folk just want a simple clear standard. The industry has now established the format for high Rez even though I am a bit sceptical as it seems just another way of extracting more bucks from your wallet and CD quality seems to be good enough in many cases. I don't feel hirez to be that killer format unless it is well mastered in that format.
However as DSD seems to be fading away the formats for experiencing digital audio or should I say methods of playback still seem to be diverse. We still have CD, CD via an off board DAC, ripped files either delivered from a laptop, streamer or media player with NAS and the options for a very convoluted path in order to eliminate noise and jitter. And finally we have direct streaming with the likes of Spotify, Tidal etc depending on the quality you want to play back.
I wonder if in a few years time if we will still have as many ways of playing back digital music or will we just be left streaming?
Spectral Morn
01-12-2018, 12:03
Interesting reading this post a few years on now the digital formats seemed to have settled down with 24 bit hirez now accepted as the standard for high resolution digital audio and DSD seems to have just faded away and died, probably due to lack of available material and folk just want a simple clear standard. The industry has now established the format for high Rez even though I am a bit sceptical as it seems just another way of extracting more bucks from your wallet and CD quality seems to be good enough in many cases. I don't feel hirez to be that killer format unless it is well mastered in that format.
However as DSD seems to be fading away the formats for experiencing digital audio or should I say methods of playback still seem to be diverse. We still have CD, CD via an off board DAC, ripped files either delivered from a laptop, streamer or media player with NAS and the options for a very convoluted path in order to eliminate noise and jitter. And finally we have direct streaming with the likes of Spotify, Tidal etc depending on the quality you want to play back.
I wonder if in a few years time if we will still have as many ways of playing back digital music or will we just be left streaming?
I would hardly say settled MQA is the new bully boy on the block to try and persuade us to buy new copies of what we have already.
DSD was also a niche thing, but I think its still there just not in big demand. Physical SACDs still sell well in Japan, and many of the high end Japanese equipment makers still offer SACD players. I don't know how big file playing is in Japan.
SACD may now only be a very tiny slice of the digital action. MQA is still very niche and as a percentage of all digital formats used is probably infinitesimal compared with Spotify or even vanilla 24 bit downloads or streaming. I bet CD usage is still vast compared to the amount of folk who use MQA?
The only two formats that are still growing are Streaming and vinyl. Streaming will undoubtedly be the digital format of choice for most people with all other digital media falling way behind. Only audiophiles will bother with Tidal as Spotify and Apple Music become the most dominant forces in digital music consumption and most people will only listen to something near MP3 quality.
Would be interesting to see a usage breakdown of digital music by category ie CD, 24bit, SACD, MQA, DSD, MP3, discs, files and streamed.
hornucopia
01-12-2018, 15:49
I'd guess the high % of the population don't know or care about 24 bit, lossless, etc.
Just us hi-fi fans.
"The Industry' is just there to sell product, not music. Streaming is so prevalent that, for example, new Mini's don't have a CD player.
It's all about you paying for stuff/'Smart' (not the users!) phones.
In my (limited)research, Lossless via HDD/dac seems 'the best', though Vinyl hardcore will disagree i know.
Also depends on quality of the recording.
Yes lossless via SSD on a server via DAC is the best I have heard
So many factors in this...and so much Fear Uncertainty and Doubt (FUD)
Greater than 16/44 wise, the “all you’ll ever need” sweet spot for me was 24/96...and the improvement was there.
Vinyl on paper has much lower numbers than this...so so it should should poor in comparison?
Back in the day I blindly upsampled everything to 32/384 in Jriver and output to my Dac and called it done.
Experience (not to mention equipment changes) led to so many WTF moments with vinyl playback (and with improving equipment well mastered MP3 320K playback) of the same recordings that taught me it’s not about the numbers, because EVERYTHING matters.
Got a botched money grabbing ‘remaster”, well guess what, it won’t be as good as the original.
Nowadays, cos of my Dac, I do DSD with everything digital, but not specifically to do with the format itself (Its complicated)
But all that said, for me...
some of my favourite recordings are on MP3 320K, some on double rate DSD and others on vinyl.
You can’t legislate for this stuff....seriously.
I just enjoy the music nowadays, whatever form it comes in.
Any tampering, Upsampling or general arseing around in the digital realm will “mess” the sound up, keep it untouched!
Big hullabaloo about classic artists of the 50’s and 60’s and the ‘pure’ recording technique ‘back in the day’, what it says is at the point of recording DO IT RIGHT, then leave it alone!
A guy I’ve spoke with David Elias does his own recordings in DSD, if your DACs compatible download something and see how good it should be.
https://exasound.com/Blog/tabid/74/articleType/ArticleView/articleId/103/David-Elias--Free-DSD-Downloads-in-Stereo-and-51.aspx
A lot of sensible true words there Gaz. I am sure DSD can sound amazing but really there is so little available it is hardly worthwhile. I read a review recently on a great new DAC and the reviewer said that DACs were being marketed with the capability of playing every digital format out there but in most cases they will never be used to convert anything other than 16 bit and some 24 bit material. It was a bit like the THD thing with Japanese apps in the 80s -so what?
Like you said everything does matter and you have to get so many ducks in a row to make digital sound really good.
Vinyl as we know just either does it or it doesn't depending on the production and mastering but when it does it makes you wonder if digital will ever deliver on the promise to be the perfect sound forever!
I have been encouraged and impressed over the last 20 years how things have moved forward with digital and we have now learnt it's not just about the bits but how those bits are delivered and kept in tact without any loss or corruption on route. I believe it is not so much about the digital recording but everything that happens to it afterwards before
eventual playback in our own system.
brian2957
03-12-2018, 08:30
I've heard Gary's DAC a few times , in my own system and in his . In my opinion it's stunning :)
The 24/96 files often have different masterings than CD. It is a real minefield trying to nail down true differences. Actually it’s easier in my system to, because 96kHz files mean no NOS treble rolloff, but that’s an unfair comparison.
IMO the DAC itself is more important than the format. The production and the mastering does rule though. Loads of shitty 24/96 remasters that just don’t sound as natural as the original 16/44.1s, for example.
I have been encouraged and impressed over the last 20 years how things have moved forward with digital and we have now learnt it's not just about the bits but how those bits are delivered and kept in tact without any loss or corruption on route. I believe it is not so much about the digital recording but everything that happens to it afterwards before
eventual playback in our own system.
I hold the opposite view which is that next to no progress has been made as there was very little room to improve to begin with. Delivering the bits is trivially easy which is why what I am writing now will appear on your computer screen without any missing letters. The difference being that you are not listening but reading. The problem isn't getting the digits from A to B it's the noise that is transferred down the wire along with the signal. Doesn't matter if the signal originates from a lap top, a CD player or a dedicated streamer, if care has not been taken to eliminate the transfer of noise what comes out of the speakers will have less clarity. The specific digital format used makes no difference to this.
When I went over to Ollies the other week we had a listen to a file played from his lap-top. Just a standard PC hooked up to a DAC. Sounded okay. Then we had a listen to the same recording on his SP10 turntable set up - clearly better.
I had brought him over a CD player, nothing fancy, a Sony CDP791E which cost about £250 back in 1995. (About £500 in today's money). So not bargain basement but not high end either. This was clearly better than his lap-top/DAC combo. But why would this be? They are both transferring the signal intact. The difference is that the lap top was not optimised to deliver the signal with a minimum amount of noise, and the CD player has been (within the limits of what can be done in that respect at the price point).
In short, it's the power supplies and how much noise they transmit down that 'digital cable' along with all the noughts and ones that degrades the sound quality. It's nothing to do with the digital side of things at all.
One last point - in the o/p Jim describes the comparison of Love Over Gold between 16/44.1 and 24/196. But Love Over Gold is an analogue recording. So both maximum possible frequency response and dynamic range is limited by the analogue tape machines used to record it. None of which can achieve a 16 bit dynamic range or are flat to 22 Khz in frequency response. In short it is impossible for there to be a 'hi rez' version of this album so any differences heard would not be down to the format. It's simply impossible. You can't reproduce what is not there to begin with.
Ironically it is the punter's obsession with illusory magic bullet solutions like 'high rez' that is actually holding back better, cheaper, digital sound. The market gets what the market wants, not what it actually needs. Look at the scientific studies, the blind tests. When they don't know which they are listening to, no-one can tell the difference between 16/44.1 and higher sampling or bit rates! Doesn't that ring alarm bells with anyone?
I hold the opposite view which is that next to no progress has been made as there was very little room to improve to begin with. Delivering the bits is trivially easy which is why what I am writing now will appear on your computer screen without any missing letters. The difference being that you are not listening but reading. The problem isn't getting the digits from A to B it's the noise that is transferred down the wire along with the signal. Doesn't matter if the signal originates from a lap top, a CD player or a dedicated streamer, if care has not been taken to eliminate the transfer of noise what comes out of the speakers will have less clarity. The specific digital format used makes no difference to this.
When I went over to Ollies the other week we had a listen to a file played from his lap-top. Just a standard PC hooked up to a DAC. Sounded okay. Then we had a listen to the same recording on his SP10 turntable set up - clearly better.
I had brought him over a CD player, nothing fancy, a Sony CDP791E which cost about £250 back in 1995. (About £500 in today's money). So not bargain basement but not high end either. This was clearly better than his lap-top/DAC combo. But why would this be? They are both transferring the signal intact. The difference is that the lap top was not optimised to deliver the signal with a minimum amount of noise, and the CD player has been (within the limits of what can be done in that respect at the price point).
In short, it's the power supplies and how much noise they transmit down that 'digital cable' along with all the noughts and ones that degrades the sound quality. It's nothing to do with the digital side of things at all.
One last point - in the o/p Jim describes the comparison of Love Over Gold between 16/44.1 and 24/196. But Love Over Gold is an analogue recording. So both maximum possible frequency response and dynamic range is limited by the analogue tape machines used to record it. None of which can achieve a 16 bit dynamic range or are flat to 22 Khz in frequency response. In short it is impossible for there to be a 'hi rez' version of this album so any differences heard would not be down to the format. It's simply impossible. You can't reproduce what is not there to begin with.
Ironically it is the punter's obsession with illusory magic bullet solutions like 'high rez' that is actually holding back better, cheaper, digital sound. The market gets what the market wants, not what it actually needs. Look at the scientific studies, the blind tests. When they don't know which they are listening to, no-one can tell the difference between 16/44.1 and higher sampling or bit rates! Doesn't that ring alarm bells with anyone?
Martin, you seem a bit confused as your opposite view is actually what i was suggesting. I would also suggest CD players are not optimised to deliver the 0 and 1s without distortion/corruption either, that is why most digital audiophiles have moved too file based audio which is best delivered from an optimised server. Lap tops are a no no in my book and indeed some CDP may sound better but in my experience they cannot compete against a good server set up. Obviously this is not something you have tried yet?
Martin, you seem a bit confused as your opposite view is actually what i was suggesting. I would also suggest CD players are not optimised to deliver the 0 and 1s without distortion/corruption either, that is why most digital audiophiles have moved too file based audio which is best delivered from an optimised server. Lap tops are a no no in my book and indeed some CDP may sound better but in my experience they cannot compete against a good server set up. Obviously this is not something you have tried yet?
Not in my own home. But in principle there should really be no difference between a notional perfect cd player and a notional perfect file based system. They face exactly the same problem, the only difference between them is how the data is stored and read from that storage. And that will not affect sound quality at all.
In practice it will depend on the individual implementation and that tends to vary wildly. Hence why for every person that says file is inherently superior you will find a person who has gone back to CD because they reckon that is inherently superior. All bollocks of course but people are very much swayed by what they hear without doing any proper analysis as to why they are hearing it and then jumping to the wrong (usually over-simplistic) conclusions.
A lot of sensible true words there Gaz. I am sure DSD can sound amazing but really there is so little available it is hardly worthwhile. I read a review recently on a great new DAC and the reviewer said that DACs were being marketed with the capability of playing every digital format out there but in most cases they will never be used to convert anything other than 16 bit and some 24 bit material. It was a bit like the THD thing with Japanese apps in the 80s -so what?
Like you said everything does matter and you have to get so many ducks in a row to make digital sound really good.
Vinyl as we know just either does it or it doesn't depending on the production and mastering but when it does it makes you wonder if digital will ever deliver on the promise to be the perfect sound forever!
I have been encouraged and impressed over the last 20 years how things have moved forward with digital and we have now learnt it's not just about the bits but how those bits are delivered and kept in tact without any loss or corruption on route. I believe it is not so much about the digital recording but everything that happens to it afterwards before
eventual playback in our own system.
Absolutely James,
so many factors can get in the way of that elusive original quality.
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.