So an end of an era today as my Meridian 565 left to a good new home. This was the lynchpin to my first attempts at computer high fidelity, as well as one of my most aspirational purchases from when I selling hifi, I still remember it being launched. It left the house for £132, originally retailing at about £3000. Happy to see it go to someone who will enjoy it. I think the recent dive in residuals largely due to it not supporting the higher resolution formats (tops out at 48khz). Is that a worthwhile devaluation for that additional resolution? When I replaced it with a 568 I did that off the back of CD Music/BluRay Films based performance - it's clearly better at both. But would it be justified just for the 24/96 capability? If you don't have High Res Audio, are you don't are you missing something special?
I think today I've got to the bottom of it.
So my last test was with the Beatles White Album, and in Blind ABX listening I couldn't reliably tell apart the CD and Pseudo HD (24bit 48khz) audio versions, but that is a remaster of a 40 year old recording. Before that I looked at a Bad CD master, against a good HD and Vinyl master, and found the later very close, and the CD very noticably worse, but mostly due to the dynamic range and not the technology. Today I really wanted to rerun the test with something modern, which was recorded with high definition format and audience in mind, and that the internet really rated the high definition version of.
http://www.computeraudiophile.com/ of great help with picking a record, and that record turned out to be the Blue Coast Collection, nominated in this post by Jud.
I bought it on the Direct Stream Digital (DSD) DFF download ($50) from the nice fast site Blue Coast has.. DSD is the format used on Super Audio CD format (SACD) and is a complex single bit process sampling at 64 times the resolution of CD. It's actually quite difficult to play like that so JRiver Media Centre turns it back into PCM audio 64bit 352.8khz which gives you an idea how much more data and resolution there is in the file. The question being whether I can notice that difference. The highest JRiver or my Soundcard can output is 24bit/192khz, and the Meridian 568 DAC can only handle is 24/96. Ok so that's a lot of numbers.
Essentially though what I've chosen is tracks are recorded and mastered with a target resolution far in excess of what my system can possibly access, and I'm converting down to the highest the system can handle, and to CD quality 16/44.1 as a comparison.
Conversions were done using Korg Audio Gate which is very effective software, but is generating Korg a lot of seriously bad Karma as unless you have some Korg gear, it forces you to tweet each time you use it! I would have quite happily paid them some small amount of money for a licence, but there wasn't an option. Really not the way to do social business.
JRiver can do very good conversion on the fly and sorts the levels out, however I did a 24/96 and 16/44.1 conversion in audiogate to ensure that the levels were identical between the two(-3.5DB to original DSD which can go over 0DB) and making sure JRiver could play them native so there was an absolutely fair back to back without one having a further DSP conversion.
The plan then was to do a set of serious ABX blind tests, and compare how much better 24/96 was than 16/44.1. But before I come to what actually happened with those results, let me talk briefly about the music
Blue Coast Collection (E.S.E Sessions)This is the single best sounding album I've ever heard on my setup at home. The clarity is awesome, the stereo and positioning totally sharp and locked, it's amazingly crisp, beautifully clear, hugely powerful without distortion, with a great dynamic range. The music is all recorded without headphones, overdubs or digital affects, it's acoustic guitar of different types, banjo and singing. It's hugely musical throughout with a fantastic sense of performance, and so transparent it's like it's happening in your living room.
So if nothing else, I have found a new set of reference tracks. If anyone comes round and says: "so all this kit, and xyz khz is it worth it?" I shall pull this out saying: "well I've got this amazing ultra high resolution DSD file, have a listen" and then blow them away with it. However why it sounds so good bears a little further examination.
Test Method and Equipment
As per original Muse test
This is where it all starts going off script. Both the WAV (16/44.1) and FLAC (24/96) sound fantastic. I went back and forward a lot listening overall. Then I went back and forward dissecting tiny sections and listening to specific parts, and I get to the point where I convince myself, that I can just hear something little better in the HD version, in the fret noise, the gliss on the guitar strings, in the overall room abience, and distinctions between the vocal harmonies. (I was particularly listening to the first track - Looking for a home).
I then go into blind tests - I utterly can't back up a difference. Once again, I'd be as good just randomly guessing which is which. I haven't included the results sets, as they are all junk. I'd find myself listening, wowed, convinced it must be the HD version, and it's the WAV version. Or vice versa, thinking that guitar gliss didn't sound as a "Real" on this WAV version and low and behold it's the FLAC.
End result - I can't tell the difference in any meaningfully repeatable way.
So I'm not saying, that I'm the last word in audiophile listener, nor that there isn't a difference. Just that as really enthusiastic amateur and with a long relationship with testing and my equipment, I couldn't spot anything that was worth spending the extra money on the higher resolution file.
When I was selling HiFi and multiple times since I've done so many back to back tests, and always been comfortable will finding some meaningful difference and normally clear preference. Especially with speakers I've found the tone is often clearly and obviously different, especially moving between different cabinet designs, tweeter types, or drive configurations.
But so often even if doing a relatively strict comparison you're changing a lot more than you think you are. Perhaps you don't have all the same cables between two sources, the speaker positioning is different if speakers are set up in parallel, or probably different if moving each each time, your positioning is slightly different after getting up and changing things, the compatibility of the equipment is different between the pieces being changed, or just the tone of the equipment is different. As an example if you're going between a SACD player and a different DVD-Audio player you're definitely comparing a lot more than just the format.
But in this test the only thing to switch between the two, was one press on the remote - everything is identical apart from the resolution of the file. I was really surprised, actually make that disappointed, that I couldn't reliably tell a difference. It makes me qestion how much is Psychosomatic in those other testings I did, especially where listening is minutes or longer apart. Can I really reliably remember the sort of subtle of difference that often audio tests are full of, if I can so easily convince myself of hearing a difference when aware of which is which, and so unable to distinguish them when I'm not.
So is there a silver lining - undeniably yes. The WAV sounds fantastic. I wish that my other CDs sounded this good, if they did I don't think I'd have a record deck, this sort of quality of sound is exactly what I tend to find I have to go to vinyl for, and I'm looking for my vinyl system to recreate. Apparently though it doesn't have anything to do with the resolution of the underlying technology.
HD releases of audio seem incredibly variable - full in equal measure of insensitive remasters, cynical re-releases in HD (often 5.1 downmixes) to cash in without adding any value, and then some sonic gold. There are also a lot of competing formats and technologies - nothing has fully come out the mix yet as the way-to-go. However it's a good thing, even if you don't have high res audio, as it's forcing a real look at the quality of the recordings on some great music. CD quality can often be poor. I've covered one in a detail on this blog, but started both back into vinyl and off into high def because I was frustrated by how often quality on CD was average or outright bad. Buying on vinyl is much reliably better, buying on HD very hit or miss, but has a lot more releases targeted at the audiophile and has some fantastic things available.
If you're lumbered with a CD you have bought that is bad, and are lusting after a high res release that is being raved about. Don't think you need to upgrade, the format probably isn't at fault. Get a downsample of that new release from a friend with high res - or do one yourself! Don't pirate, pay for that CD, pay for the effort of a remastered version, but if you're paying in my view you should get a version that adequately makes use of the good technology available CD. So I'd feel totally justified in replacing my Muse 2nd Law WAV from CD which I bought, with a down sample of the HD version, as the CD is just crippled by the mastering, it has nothing to do with the technology.
So it's worth a note that the 568 is up-sampling the CD to 24/96 - but to my mind this isn't relevant, if it sounds as good with the DAC guessing the additional data, you didn't need the additional data in the source, you're just relying on the quality of the DAC. Better DACs do sound better, and are probably now built to higher definition. However it's not perhaps necessary to have the higher resolution file to exclusively access that quality (although if new DACs are only tuned on high resolution material that may start to be the case)
Should I upgrade my kit just to get high resolution audio?
No, if you have a great 16/44.1 (or 48) digital system stick with it. Just taking it to 24/96 on it's own isn't going to make a big difference and it will involve changing some fairly fundamental parts of your kit with expensive new things. Spending that money elsewhere may well give you a bigger difference. If it sounded great before, it still sounds great now and good news: You can down-sample those lovely new HD remasters and they will still sound fantastic on your agree. Certainly don't upgrade your kit until you've tried a downsample of the music you'd be upgrading to get!
If I'm significantly upgrading sources anyway should I get a high resolution component?
If you are upgrading your source anyway, probably you'll end up with a high res one any way if you're spending significant money as this is where the market is moving. More material will come out for it, and getting a 24/96, or 24/192 or even 32/192 gives you future proofing. But don't think you need to throw away your CDs. If your new DAC/Source is to better, it needs to be better with those as well, not just with new files Back to back it with your current music.
If I'm getting started should I go straight for high res?
If you'll consider second hand, there are a lot of fantastic 44.1/48khz DACs out there on the market that the move to high def is driving into the market. On the same whole-system-budget of say £1000-2000 buying one of these, and matching carefully with other fantastic second hand components available, I think is going to give you a much better sound overall, than spending a lot more of the budget on a DAC and source capable of higher res nad having to compromise elsewhere. Certainly than trying to buy new High Res at this amount of money. Similarly getting hung up on 96 v 192 khz (or any other numbers) just isn't worth it. Good quality hifi kit with a lower res, will outperform cheap computer based equipment with a higher res every time. Go for quality kit.
Should I ditch my vinyl for HD?
No. Though if you have a CD Quality digital source as well, you may be pleasantly surprised by how some of the better HD releases sound down sampled to what your system is capable of. Almost not sounding like CD at all!
Is the DSD premium for Blue Coast worth it?
$20 for 16/44.1, $40 for 24/96, or $50 for DSD DFF or DSF
I can't really answer this authoritatively, so no outright recommendation here, as I didn't buy the 16/44.1 I down-sampled from the DSD. But if they take such care over the high res, I would expect they haven't messed up the down conversion.
I also didn't listen in DSD - this requires a special DAC. I might see in the future if I can borrow one to see. But there is a lot of controversy out there on whether DSD is worth it - I remain a skepti but don't have the science to back that up at the moment.
My encouragement would be to buy the 24/96, it will encourage people to record this sort of audiophile recording, and you're recognizing that with a premium. The file is future proofed to a large extent, and will be playable by a lot more things than the DSD. You also get to know that you're buying a very high quality source and utilizing the maximum affect of your kit if it is compatible with it.