I also recognized that the trumpet got a bit more bite on the remaster, but the simultaneous degradation of the piano sound and the other colorations don’t make it beneficial for me.
The Mervine seems to be the first remastered one. We’re not sure if we’re the only ones with this opinion or if no one really cares or few just bought it or no one else would be critical.
I don’t think a speaker or its suitability for monitoring and engineering work is the topic in such cases, I think the art of mixing/mastering is mainly sensible to too much or too consequential change and probably needs long experience in transferring results of one monitoring chain to a majority of setups. I bet if PSA remastered on an IRS V or a normal studio speaker and Aspens, there would have been completely different results. But I guess the goal of such engineering work is, to produce roughly the same compatible result with whatever monitoring system. Id say a remastering doesn’t have to be and should not be completely different with a new system if it wasn’t completely off before.
Overmotivated remasterings are not completely uncommon. Having read about mastering in forums like the Hoffman forum, I always wondered how small the changes made were, related to partly even quite problematic initial situations.
I don’t have the Mervine re-issue so I can’t comment on its sonics. However, I do have most of the Octave library on cd. I do not find any particular sonic signature. Neutrality, if anything.
I’ve heard much greater variance in other audiophile labels like Chesky and Reference Recordings. As for FR 30’s as monitors, neutrality rules the day here too. I own FR 30’s and have mixed on other monitors, like ATC, that are revealing, not emphasizing.
Sure, even under the greatest scrutiny, an album can sound different from the pack, but on balance, Octave has been consistent, revealing, musical.
If other labels were at this level, I would stop contending that the source is the crappy valve that we all must listen through.
I didnt understand akro questioned consistency (up to the remix). Did you find excessive highs on the old Mervine mix? I guess not from what you just wrote.
I just hear another track from the old one (Sunlight). It seems to me like a drummer’s album (and clearly more vibrating/lively/open also in the lower registers of piano and bass). The drumset is a bit more pronounced than in an absolute equal group, but that’s it in my perception.
But if the other new mixes are as different from the previous as this one, you won’t find that exact signature anymore that defined your perceived consistency.
Not sure if you followed the Analogue Productions SD Pretzel Logic dilemma. Chad Kassem seems to have “corrected” Grundman’s remaster and lowered the top end where nothing should have beeen done. Everyone else heard that and I guess there are quite some cancellations of this one. If so, that one’s a clear case of let the engineers do their job. I’m so glad I have an original, it’s just perfect., heard the digital version of the new remaster and it’s really screwed up.
Interesting, I’ve only heard the new digital remaster of Pretzel Logic, but it did sound off to me somehow, but couldn’t put my finger on it. I only do digital, but the other AP Dan SACDs have been excellent imo. Glad I have the Japan SACD of Pretzel Logic!
are all Octave DSD recordings converted to DXD (ie, 392 KHZ PCM) and then converted back to megahertz DSD (64, 128, 256) for sales as DSD, SACD, and vinyl?
That is the track Whiskey from Soundstage. I thought I did pretty well on maintaining the mix (except for the bass which always sounds best in the middle and when amplified on the stage that’s how it sounded).
Every master, every remaster is performed, at the end of the day, by ear. One way or the other, a mastering engineer is listening to the new mix and making choices as we did on the Mervine. Most that have heard it think it’s a big step forward. Clearly, others have had a different opinion.
But, at the end of the day, it’s always about how it sounds in the mixroom.
This is correct. 1-bit DSD at any sample rate cannot be volume adjusted or mixed. It must be first converted to analog, or DXD, or multibit DSD. Somehow, it must be converted to one form or another in order to be mixed or volume adjusted.
We’ve experimented a lot with the various alternatives and still find that when done correctly, in the box mixing on Pyramix remains the best option. So, unless the engineer makes a 1-bit DSD two channel recording asnd publishes that as the final output, every DSD recording you have has in one way or another been converted then back again.
Thanks Paul, yes, that’s clear, I’m sure it sounded right for you in the mixroom and when they mastered Hotel California with (as I read) the speakers under the mixing console (by the way, it sounds like that), they also had the right sound in the mix room.
Now I don’t want to imply, your mix room situation is anywhere close to a strange place, not more than any other professional mixroom, but as all mixrooms sound different and all have to produce a final recording in them, which sounds right on most systems. I guess it’s a long way in experience with the own mixroom’s (incl. equipment’s) sound, if and how it leads to most compatible and quality products.
You probably made quite some changes at the same time (room, speakers which to you and others sound very much better/different to other speakers you experienced, a different mixing process and maybe more) and you probably didn’t have a lot of time to gather experience with this situation before releasing remasterings, which “had to be” better. That’s no easy situation. But as always we’re talking about smaller differences than they read and my opinion is just one in the crowd .
And I have no better idea. You, as all others, use the best equipment you can think of (which all sounds very different from each other) for making the optimal product. How can this work. What else should you do. It seems to be whatever experience with this whole topic. I’m sure if I’d produce recordings (especially as having not a hint of the experience you have), turning the knobs, with the chain I like best, it would sound somehow weired from start to others, although great to me there.
There’s an engineering chain of a label in the market (2xHD), containing the absolutely highest audiophile home standard, which imo at least produces mixed results, good or great in most own productions, rather worse or not generally betterthan others in remasterings, although equipment wise on absolutely highest audiophile standard.
For DSD256, analog signal from microphone(s) is sampled at 11.2 megahertz (11.2 million samples per second)
Then, for mixing/mastering, the DSD256 is converted to DXD, ie PCM sampled at 352.8 kilohertz (0.3528 megahertz)
Thus, tossing 10.85 million samples per second in the down-sampling process
Then, remixing and remastering occurs comprised of adjustments, additions, and subtractions to down-sampled music, after which the modified 352.8 thousand samples per second are up-sampled to 11.2 million samples per second.
Where did the ‘new’ 10.85 million samples per second come from? Created interpolations?
Is the DSD256 we purchase mostly ‘created’ information NOT represented in the starting point of 352.8 thousand samples per second?
That’s interesting, as it’s often used as a critical vinyl argument, that in extreme cases, bass has to be centered. Didn’t know it’s better centered anyway.
But Paul, @akro is right, the mix has absolutely nothing to do with the picture. I even had to make a channel and phase test with my setup to exclude a problem on my side, but my setup as expected was correct.
On the picture the fiddle is left, then guitar, then Banjo, then bass.
On the recording, Banjo is left (he also seems to sing), then bass, then guitar, then fiddle. The whole perspective is also not centered but starts from medium left to further right.
So the most positive to say would be, you reversed channels, pushed everything a bit to the right and put the bass in the middle.
If you’re interested in the difference between DXD mixing of DSD and a way to avoid it and compare the SQ effects, try the Noqué sample files. The non DXD processed ones sound noticeably more open and airy.
But be aware, this isn’t analog vs. DXD mixing but this very special method of DSD mixing vs. DXD mixing.
I wish I understood the maths involved in rather more detail, but remember that’s 1 bit samples you are talking about, not a lot of detail in them.
When it is “down sampled” to 352, but at a bit depth of 24 bit - the information in those 24 bits comes from the information in the “tossed away” sample points.
It isn’t really “down sampling” but “cross sampling” or remodulating to a different codec.
AS I said, I do not have the maths to describe it properly, and Ted doesn’t play in this sandpit any more but he has explained this at some point or another, or someone else with equivalent insight has, on this forum.
TL;DR
The extra sample points are not thrown away, they are used to create the detail in the bit-deepened DXD.
…and then in reverse, when going from 352k/24 to 11m/1, the extra bit depth is transformed into the extra sample points, the extra sample points are not created interpolations, they are translations.
.