I ordered the SACDs yesterday. Tried to download DSF files from the Japanese e-Onkyo site, but the download is much more expensive than the physical SACD set. And the download doesn't come with any documentation. Odd decisions by either Sony or e-Onkyo! (Same with Thielemann's Bruckner cycle. The SACDs are much cheaper than DSF download.)

I've also ordered this set (cdjapan) and hopefully I'll receive the discs next week. - FYI: I asked b-sharp Berlin (the remastering studio for this edition) what kind of master files they got from Sony. They told me that Sony gave them files in PCM 24/96, which b-sharp transferred/'remastered' to DSD for SACD production.

Bruckner/Thielemann/VPO: the complete cycle is available at prestomusic and qobuz in PCM 24/96 (the genuine/original recording format) for less than € 30... - I'm really enjoying this Bruckner cycle mostly because of the sublime playing and the impressive sound. I'm quite far away from being a fan of Thielemann, there are more interesting and gripping cycles on the market for sure, but I can hardly imagine a complete cycle which is better-played and recorded. - My personal favorites from this Vienna Bruckner are: symphony in f-minor, 1st symphony (because of the late 1891 Vienna version which has a magnificent final coda).

Huh that is very interesting that they received 24/96 files from Sony considering when these recordings were made. I do know Sony was often recording at 20 bit before many other labels were, but I do wonder that 24/96 was NOT the native recording format for these Maazel Mahler recordings.

BTW I did see that the Maazel/Mahler cycle was available as a 24/96 download too. But I decided to go with SACDs. Hoping for a delivery soon (also from CDJapan).

I am also very skeptical about the file format. My guess is that the cycle was originally recorded in PCM 16/44.1. Only for the preparation of this new edition were the old files remastered or edited in some way and converted to the new format (24/96). This is very reminiscent of the procedures used by Esoteric or the more recent SACD editions from TowerRecords, in which digital recordings are "enriched" with overtones.

Ultimately, we will never know what exactly was done. So let's wait and see. I am very curious to hear whether and how the sound has changed compared to the old CDs.

I should have known better, that old PCM recordings from early digital days not really can be improved regardless whatever magical treatment or remastering they might have received. After comparing the beginning of the 2nd movement of the 1st symphony (old RBCD vs. new SACD), I can definitely say that there is some kind of a difference. The SACD sounds a bit 'tidied up', soundstage appears slightly more focused. To my ears there's no change in reverb modification. Because SACD volume is a bit higher than RBCD, comparison is (as always) not that easy. The more you raise CD's volume, the closer it gets to SACD. Overall there's really no big difference or otherwise audible surprise that may wow you. - In the end, this release fits perfectly to Esoteric's or TowerRecords SACD releases of (16/44.1) PCM recordings.

I don't regret this buy nevertheless, because I never owned the complete cycle and the extremely attractive current ratio between yen and euro made things easier.

I have to correct my previous comment a little, I was probably a little too hasty. Also, I had only compared a single track. Now a few days have passed and I've listened to a few symphonies in their entirety. As I wasn't really familiar with this cycle before, I have to admit at this point that I am extremely enthusiastic about Maazel's interpretation and, of course, the playing of the Vienna Philharmonic. I am particularly fascinated by the fact that the music is given a little more time to unfold and breathe, which often leads to incredibly beautiful sounds. This is a complete contrast to Solti's Mahler cycle, for example, which simply stomps over a lot of things.

As far as the sonic improvement of this new edition is concerned, the differences do not necessarily seem earth-shattering at first glance. However, after listening for some time, I now realise that the sound is more pleasant and - within certain limits - even more natural. Even though I am still sceptical about remastering old PCM recordings, I have to admit that Sony and b-sharp in Berlin have made a great effort with the old files. Incidentally, the booklet states that the first recordings of the cycle were made on 2-track Umatic, while the last recording (Symphony No. 8) is a 48 (!) track recording.

I really enjoy this new edition. Even if the interpretations themselves perhaps fascinate me a little more than the improved sound, I would still like to recommend it at this point.

We are living at the time of A I and ChatGPT so anything may be possible! Joking probably but will A I be able to fill in the missing data? Seems likely at some point if not for this recording.

Interesting comment, Hiredfox. That had crossed my mind, too. I'm sure AI will be capable of remastering in the near future, be it old analogue recordings or early-digital PCM files. By comparing thousands of modern high-resolution recordings with older or lower-resolution recordings, I'm sure AI could determine what the typical sonic differences are. Building on that knowledge, it could then "upgrade" these older or lower-resolution recordings. Initially, the results would surely be poor, but with time AI would improve its technique. That would of course raise all sorts of questions: how authentic would the result be? how much tampering is "musically ethical"? etc.

Food for thought...

I guess also whether it would even be financially viable, first as a development and then a production project. Look at us here, a niche in a niche in a diminutive market already, as it is. (I almost added the adjective "dying" there before the diminutive market, but let's not do that yet.)

I guess also whether it would even be financially viable, first as a development and then a production project. Look at us here, a niche in a niche in a diminutive market already, as it is. (I almost added the adjective "dying" there before the diminutive market, but let's not do that yet.)

Interesting food for thought, indeed.

On a slightly different track: Some time ago I reviewed Schubert: Symphonies 1, 3 & 7 (8) - de Vriend in which I paid attention to experiments to ‘complete’ the Symphony, including Huawei’s recent ‘smartphone’ attempt, using artificial intelligence to finish the symphony as Schubert could have done it. It was my view that it “sadly is no more than a compilation of Schubert-like tunes, missing the point, or better: Missing altogether the soul of the symphony .. ”

It can be heard here: https://consumer.huawei.com/content/dam/huawei-cbg-site/weu/common/mkt/campaign/unfinished-symphony/HUAWEI-Unfinished-Symphony.mp3 from min. 24:35 onward.

Not quite the same as discussed here. But one wonders if a computer can do more or better than what has been put into its memory or is available on the Internet.

Besides questions raised by Hiredfox and Athenaeus, one might also wonder to what extent filling in missing data by AI will lead to improved sound.

The AI question is interesting and nobody knows for sure where future will lead, but when it comes to creative things, AI isn't capable because it's not a creative thing. It can take existing information and put it into a blender, and *something* will come out. But it can't be creative in the same way the human brain can be. It can take a 16 bit digital recording and use other recordings to create... *something* to make it different, but it doesn't know if it's improving a thing. (Of course the word "improving" can mean different things to a million different people.) A human, whether Ward Marston or the people who work at Esoteric or really, anyone who remasters audio and video files, makes judgments about every second of that recording. Not all "pops" are bad. Will AI know which pop to remove? I don't see how it can unless it develops independent thought like in the movies, and we long from that if it is possible at all. Same principle would apply to AI finishing Schubert's symphony. As Adrian describes, AI took what it considered Schubert-like themes (which were probably tagged as Schubert-like by a human when cataloged) and gave us a soup of what it thinks Schubert would have done. Of course no AI - or any human - could have any possible idea what Schubert would actually have done!

But about this Vienna/Maazel cycle - I'm enjoying the performances enormously. I think the 2nd Symphony is one of the best performances I've ever heard. And the 4th is fantastic also. (This 4th with Battle was the first Mahler recording I ever bought... on cassette!) The sound...well, who knows. Who knows what it would have sounded like if it was recorded in HD. I don't have enough of these recordings in 16 bit to compare. But 2 and 4 sound fine, warm, but maybe the sound isn't as open as good 24 bit/DSD recordings can be. It's not boxy, but it just doesn't seem to bloom as much as I'd personally like. However, I am not regretting the purchase and am looking forward to listening to the rest.

What was recently done with John Lennons home made tape could possibly clean up say Caruso as well. It’s not AI, it’s what a capable sound engineer can do with the help of AI.