Penteo surround demo

[Cai Campbell]

Sure I have no problem with that. What format do you want the output in? DVD-Audio okay? It might take a couple weeks since I do have stuff to do. The output would be 5.1 channel 88.2/16, since that's a simple math upsample of the original mix with no dithering.

DVD-A would be great, but can you also do a DTS CD? The bulk of my audio system is in storage at the moment and right now DTS CD is the only surround format I can listen to.

 

[phydo]

Quadfather:

Well said!

Justin

 

[penteosurround]

DVD-A would be great, but can you also do a DTS CD? The bulk of my audio system is in storage at the moment and right now DTS CD is the only surround format I can listen to.

I can, sure. it requires a separate mastering, so I'd like to do one or the other. I'm actually surprised that you would accept a compressed version! But sure, DTS is fine. Give me a couple weeks.

-John

 

[penteosurround]

Actually I wasn't going to weigh in on this, but the discussion has gotten rather intense. It seems to me that the Penteo process would be very valuable to surround enthusiasts, but could also be damaging depending on how it was used. My concern is that it would be used in place of a true surround mix when it would be possible to create one, or where one already exists. Will the radio stations broadcast the Penteo versions and not the real surround versions where one is availlable? And when there is a good quality quad mix, will that be used instead?

Who knows. Right now the surround radio format isn't even in its infancy, it's in its gestation. If you want to read all about it, check out http://www.nrscstandards.org/DRB/misciboc.asp. There are two camps: the matrix people (Dolby, SRS, and Neural) who are all proposing matrix synthesized surround ideas (which is what I think you guys are arguing against!) vs. me and Telos/Fraunhofer who are doing true panorama parsing and encoding it in true multichannel MPEG4 for transmission (Fraunhofer invented MP3 and MP4) and keeping true multichannel transmission. The matrix methods (for upmixing) are pretty much the same as playing back stereo through your home stereo in DPLII modes, so no prep-work or remastering is necessary. It's truly synthesized surround.

Penteo on the other hand, as I've said, is nothing synthetic; it's a panorama-based "crossover network", plus I actually remaster each song, taking great care to optimize each master prior to panorama separation. And with Penteo 1) the mix downmixes back to the original stereo, since all we're doing is separating out the original pan positions; 2) all 8 channels are in sync and in-phase (5.1+stereo=8 channels), which minimizes the data that has to be transmitted for the rear channels since those can be sent out as difference information, and 3) I've written all the code to automatically import the AudioVault library information and code it into the 8-channel Broadcast Wave File Format which is reverse compatible with the existing hardware that's in use in most radio stations for music library database.

Please realize that the operation of a surround radio station in any discrete format requires simultaneously playing two versions of every song: the stereo for the stereo listeners, and the surround version for the surround listeners. It's actually very easy with Penteo because the file format for broadcast is 8 channels already in sync. Yes, you could use the 5.1 remix or the quad mix for the surround channels.

So yes, I can take the new remixes (and the original quads) and incorporate them into the AudioVault 8-channel format, but it requires hand-syncing them up to the original stereo mixes so that they can be encapsulated into a single file, since the 5.1 remix or quad mix might sound awful when downmixed into stereo, since it was never intended to be downmixed. You can see how difficult this is for a library of thousands of cuts. That's one of the reasons that Penteo is so perfect for large-scale processing of libraries. That said, Penteo is still a very time-consuming process, since each mastering error, azimuth error, original splice and EQ match must be found, and fixed up if necessary. All these things take time.

I realize that some songs especially those older ones that were recorded before multitrack recording was widely used will never see the light of day as surround songs except for the use of this or a similar process, and for this, it is extremely valuable. As far as arguements go, as to whether the result is true to the stereo mix, it's not germane, for we are in it for surround, having declared that mere stereo isn't good enough. At least that is the case for me. I just want to make sure this process or similar ones are not used as a replacement for the real thing.

That's where I get confused with terminology. By "the real thing" do you mean the original quad? Do you mean a new 5.1 remix? The "real thing", for the most part, was the original stereo mix that became famous and burned into our collective memories.

I'm not in any way trying to stifle surround remixes as an art form unto themself. What I am trying to do is to showcase how incredibly good and complex the original stereo mixes are, and by simply separating out the pan positions with a fine-toothed-comb, one can marvel at how many tracks - how many hidden performances - have been there in those mixes all these years.

-John

 

[Cai Campbell]

I'm actually surprised that you would accept a compressed version!

Believe me, it is not my preference but is a necessity if I want to be able to listen any time soon. I'm in the process of building a new house and virtually everything I own is in storage, but I am able to listen to DTS.

 

[ssully]

I can only assume that you are trying to address the stereophile nay-sayers who poo-poo surround sound because it deters from the original stereo mix, that "original work of art" you refer to. Do you REALLY think these myopic stereoheads will buy into a fake surround presentation just because it is based on their beloved stereo master? I think not!

I think so. In fact, one of the thing that often bothers me about surrond remixes is that parts of the original mix are missing; it's why I often prefer a DPLII 'remix' , though it's far from discrete. The Penteo process sounds like a surround extraction process that keeps all the original elements, but produces a more discrete surround mix.

 

[ssully]

I do have some questions for Mr. Wheeler though -- how are you getting access to all those stereo master tapes?? THat's an amazing array of tracks on your list -- are you working from original two-track masters for all of them?

And do you ever compare the master to an old LP, to make sure that no EQ/level change/fading was done during cutting, or parts added? I can think of a few tracks where the CD (re)masters are different from the LPs, because these 'last-minute' changes weren't taken into account (e.g., Yes 'Drama' where an LP fadein is not replicated, Jethro Tull 'Heavy Horses' where some string parts have gone missing; also, with the recent LP remastering of Yes' Fragile, it has been noted by Steve Hoffman that the original tapes have to be re-EQ'd to get close to the sound of the original LP release; the first CD release of the album in the 80's seems not to have done this, and was quite dull-sounding compared to the LP).

 

[jaybird100]

I think so. In fact, one of the thing that often bothers me about surrond remixes is that parts of the original mix are missing; it's why I often prefer a DPLII 'remix' , though it's far from discrete. The Penteo process sounds like a surround extraction process that keeps all the original elements, but produces a more discrete surround mix.

OK, here's a question that I don't think has been addressed here. If, for some reason, someone wanted to listen to a Penteo-processed recording in regular stereo, how faithful to the original stereo mix would the Penteo version, collapsed into two-channel, be to the original mix?

 

[Bob Romano]

I think if you look through the three pages (or so) here you will find that that is exactly the point of the Penteo system. It does nothing to add or subtract from the original stereo mix but takes things within the stereo mix and spreads them out (for ease of explanation). When you put all the parts back together your stereo mix is intact.

 

[jaybird100]

I think if you look through the three pages (or so) here you will find that that is exactly the point of the Penteo system. It does nothing to add or subtract from the original stereo mix but takes things within the stereo mix and spreads them out (for ease of explanation). When you put all the parts back together your stereo mix is intact.

I think I'd actually have to hear them compared side by side to believe the Penteo processing doesn't change something in the original mix when played back through two speakers. I may not be from Missouri, but I'm skeptical enough to say "show me".

 

[Old Quad Guy]

I do have some questions for Mr. Wheeler though -- how are you getting access to all those stereo master tapes?? THat's an amazing array of tracks on your list -- are you working from original two-track masters for all of them?

That's the part that does not add up. He told us he's working to create surround from the stereo master of "Seals and Crofts Greatest Hits" when the quadraphonic mixes still exist for "Seals and Crofts". All I'm asking for - since he is already working with one of the major record companies with surround - it would be greatly appreciated if he could help us get some of the original Quadraphonic mixes rereleased in some format. If HD Radio gets off the ground and the technology is there for all types of surround - that would offer an array of surround broadcasting options.

What Penteo is or is not and how does it sound (are there artifacts in the sound process for instance) is still under consideration. If Penteo can get HD Radio going using a true multichannel transmission that could also work for 5.1 SACD and DVD-A Broadcasting (including 4.0 Quadraphonic) I would think that would be good.

The new thread with new DTS sound samples:

https://www.quadraphonicquad.com/forums/showthread.php?t=7708&highlight=penteO

 

[penteosurround]

I do have some questions for Mr. Wheeler though -- how are you getting access to all those stereo master tapes?? THat's an amazing array of tracks on your list -- are you working from original two-track masters for all of them?

The classic rock cuts that I am working from are from the collection that CBS/WZLX is using (and has been using) for its owned-and-operated radio station(s). There are 1500 cuts overall. Most are from CDs, though some are taken from other sources. That's why step 1 of the process is master correction; no master I've worked with from an analog source is absolutely perfect; there are always some azimuth errors that tend to wobble at every tape splice which would never have been caught in a traditional mastering session. Also the left/right balance is usually off by a few tenths of a db, sometimes as much as 3db. Usually that's because the mix on the master tape doesn't match the tones at the head of the reel, which happens a lot.

In addition, I have to maintain the database and "CART" information for each cut, which the automation system (operator assist) uses to count down the intro, etc. All this info I then multiplex into an 8-channel 44.1 WAVE_FORMAT_EXTENSIBLE .wav file. for radio station use.

-John

 

[penteosurround]

I think I'd actually have to hear them compared side by side to believe the Penteo processing doesn't change something in the original mix when played back through two speakers. I may not be from Missouri, but I'm skeptical enough to say "show me".

Not only is it intact, but if you mix the downmixed Penteo version 180 degrees out of phase with the original stereo mix, you get silence. They perfectly match and cancel each other out.

 

[ssully]

The classic rock cuts that I am working from are from the collection that CBS/WZLX is using (and has been using) for its owned-and-operated radio station(s). There are 1500 cuts overall. Most are from CDs, though some are taken from other sources. That's why step 1 of the process is master correction; no master I've worked with from an analog source is absolutely perfect; there are always some azimuth errors that tend to wobble at every tape splice which would never have been caught in a traditional mastering session. Also the left/right balance is usually off by a few tenths of a db, sometimes as much as 3db. Usually that's because the mix on the master tape doesn't match the tones at the head of the reel, which happens a lot.

Hmm, color me skeptical regarding someone's ability to 'correct' a 'master' that's a commercial CD that's already been through at least TWO masterings (the one creating the original master tape, and the one creating the digital master). So far you seem to aim to correct setup problems with the original tape recording, but how do you 'correct' tasteless EQ moves and over-application of digital compression or noise-reduction to the CD master, especially not knowing exactly what was done in the first place? I'm really not sure how these would be marketed. 'Mastered from a 1997 CD remaster'? I'm not sure that will grab the audio-for-audio's sake market.

 

[spenceo]

It's my understanding that these cuts that penteo makes are to fill in the gaps of a MCH HD radio station's play list (WZLX-FM, in Boston). It seem that the HD radio station uses as many true 5.1 and 4.0 mixes that they can gain rights too. I don't believe any of these will be for sale in some form. I personally see this as a great thing for surround, another attempt to get surround to to the "average Joe".

Spence

 

[penteosurround]

Hmm, color me skeptical regarding someone's ability to 'correct' a 'master' that's a commercial CD that's already been through at least TWO masterings (the one creating the original master tape, and the one creating the digital master).

Spenceo is correct; these cuts that I'm working on are not for consumer release, they are for broadcast. And as far as mastering correction, no compression or anything has been applied, these are digital clones of the original 44.1 mastering session at the mastering house which prepped the CD. And there's only one "mastering", that's the transfer from analog to digital. Everything after that are clones of that master; none of what I'm working on has had any compression--or for that matter ever even been audio--since being mixed to 2-track. It's just a bunch of zeros and ones. The correction that I'm doing is for the ever-so-slight physical head misalignment when playing back the master tape, or else a very slight (10ths of db to 2-3 db) mis-balanced left/right level. These are errors that have always been there in the CD, but it never made any difference because they are inaudible errors. The Penteo process requires perfect time and amplitude alignment, so anything that is even slightly off shows up clearly, so it has to be micro-corrected.

-John

 

[ssully]

Spenceo is correct; these cuts that I'm working on are not for consumer release, they are for broadcast. And as far as mastering correction, no compression or anything has been applied, these are digital clones of the original 44.1 mastering session at the mastering house which prepped the CD. And there's only one "mastering", that's the transfer from analog to digital.

I don't question that a CD is a digital clone of its digital master. And I realize your current project is for a radio station, but you indicated you intended to offer this service to record companies, though the market isn't too good for surround releases right now.

Anyway, the original analog master tape is a "master" too, so if it's remastered to digital, that's two masters, and there's no guarantee the digital master is just a flat digital copy of the original MT.... and surely you're aware that many albums have been rereleased several times on CDs, with different mastering on each? To me those seem like complicating issues, unless you have managed to find sources that are authentic digital 'clones' of the original master tapes, without any tweaking done after the digital transfer.

Everything after that are clones of that master; none of what I'm working on has had any compression--or for that matter ever even been audio--since being mixed to 2-track.

This is the part I'm not understanding. Where did the radio station get these digital masters of tracks by Led Zeppelin, Yes, etc (which were Atlantic, not CBS artists)? Did they have someone go back to the original two-track analog master tapes and make digital transfers just for the station? I find that really hard to believe.

So, if in fact your are actually working from tracks sourced from commercial CDs, especially ones mastered in the last 15 years or so, there's quite a good chance that digital compression and/or noise reduction and/or EQ were applied -- things that aren't on the original master tapes.

It's just a bunch of zeros and ones. The correction that I'm doing is for the ever-so-slight physical head misalignment when playing back the master tape, or else a very slight (10ths of db to 2-3 db) mis-balanced left/right level. These are errors that have always been there in the CD, but it never made any difference because they are inaudible errors. The Penteo process requires perfect time and amplitude alignment, so anything that is even slightly off shows up clearly, so it has to be micro-corrected.

-John

I think you're missing my point, or else I'm totally misundertanding what you are working with. If you're not working from original two-track analog masters of these albums, or flat digital clones thereof, then there are all sorts of questions which arise regarding 'correction'.

 

[jaybird100]

Why has this thread taken the tone of ganging up on John? He's trying to further our cause with a system to create surround sound where there was none. I agree it's not intended to replace true discrete 5.1 mixes, but if it gives us what we crave in one form or another, I'm all for it.

 

[penteosurround]

Why has this thread taken the tone of ganging up on John? He's trying to further our cause with a system to create surround sound where there was none. I agree it's not intended to replace true discrete 5.1 mixes, but if it gives us what we crave in one form or another, I'm all for it.

Oh I don't mind, I'm used to it. Ssully, what I'm working with are files - files that are 44.1 clones from CDs or 2-track transfers from special radio mixes/dubs/edits, which came from the same mixers and sessions that made the albums; they've just never been commercially released, they were made as "radio mixes" by the original mixer.

Honestly I don't know where all of them have come from genealogically, and since they would never be used in a commercially released product by anyone, it doesn't really much matter. For our prototype experiment in Surround Radio, they sound incredible.

As far as how the sausage is made: normally when a bunch of hits, for example, are re-released as compilations, the original 44.1 transfer (which is on a F1 or PCM701 3/4" U-Matic video tape if it's from the 80s or before ((the whole reason that the 44.1 rate was used in the first place was because of its relationship to the videotape's horizontal sync frequency))) is used and just transfered straight to a workstation these days. Almost no one EVER gets the 2-track out of underground storage; it's just too fragile to handle (and valuable, so the liability has to be taken into consideration) and has to be baked in an oven anyway. Besides, some of the tapes have way deteriorated from the mid 1980s when the original 44.1 transfers were done to F1.

Usually when I load a file up into my workstation, you can see if any dynamics processing has been done - let alone hear it. I can honestly say that out of the original 42 cuts that I have done for the 'ZLX project, only one had had compression at the post-mix level, (you can tell because the noise floor pumps or else the waveforms are flattened at the peaks) so we had it re-transfered from a different source. We're all audiophiles here. You can tell just by looking at the waveforms; the uncompressed cuts don't have average levels that exceed more than -10 or so, with peaks that jump out 10-15db higher than the average level, at least. Sometimes you realize how much "louder" you could make the song overall if you just clipped a peak here and there (like a haircut with only 2 hairs sticking up) but since you want to remain faithful to the transfer, you don't touch a thing. Of course once they're on the air, they go through Optimods that clip everything anyway.

You'd actually be amazed at how many CDs are out there that truly are just straight flat transfers out of the 2-track analog playback. Mixes that were optimized for AM radio still sound "peaky" in the 5-8kHz range, which they are. Of course in some cases, that "sound" is what made them hits in the first place.

-John

 

[Old Quad Guy]

If Penteo can get HD Radio off the ground using MPEG 4 in a true multi-channel transmission (4.0/5.1 SACD/DVD-A Broadcasting) instead of the matrix options - then finally I'll get my vintage 24/7 discrete "Quadraphonic only" HD Radio station.

So how do we get HD Radio off the ground for our Quadraphonic station?

It's not going to be easy with so many competing broadcasting formats.
Surround sound educational document (PDF Doc)
http://nrscstandards.org/DRB/SSATG report final.pdf

From the NRSC site:
http://www.nrscstandards.org/DRB/misciboc.asp