Article: Trent Reznor: This One's On Me

[pichonCalavera]

Woohoo!, free music!<br />

<br />

I think Nine Inch Nails has really nailed the download releases, since they go great lengths to making it "an event" and they publish many digital format files and release "album artwork".<br />

<br />

I myself have my music in FLAC format, so seeing FLAC downloads is great stuff (altough they are distributed in .torrent format, but that is no issue for me)<br />

<br />

[poop]

Yeah I'm not really a fan of torrents - in fact I had to install Azureus just to download the album.<br />

<br />

Have read on a few sites that the 24/96 Wav version is more compressed than the FLAC version. I have only compared through DAC1 USB into Senn HD650 at this stage, and can't hear a difference between the WAV and ALAC versions. I expected the difference to slap me in the face so am a little surprised... Any comments/info from anyone able to test this?<br />

<br />

Just adding a bit more info from another site regarding the encoding of 24/96:<br />

<br />

"What I can see that it's essentially a 16-bit signal converted back to 24-bit. I don't know how they converted it to 16 bits, it was probably dithered, but then it's dithered 16-bit data with additional 8 bits of nothing"<br />

<br />

Seems to echo others findings too...

[The Computer Audiophile]

Hey Poo - I downloaded the 24/96 version but haven't had enough time to really listen to it. Now I will have to download the other versions and let my ears make the decision for me. <br />

<br />

Traditionally I haven't been a fan of torrents either because of the bad connotation the word torrent has. 99% of the time torrents are used for illegal file sharing. It is fabulous technology and I'm very happy to see it put to use for this distribution.

[poop]

http://www.hydrogenaudio.org/forums/index.php?showtopic=63086&st=50<br />

<br />

This is from Hydrogen Audio - someone has captured waveforms that illustrate the clipping...<br />

<br />

Surprisingly the 96/24 version appears to be more compressed and clipped. Something must have gone terribly wrong.<br />

<br />

Here is an example - Amplitude statistics and waveform screenshots from Adobe Audition:<br />

<br />

44.1 kHz 16-bit, Track: 10 Demon Seed, passage: 3 min 50 s - 4 min 00 s<br />

<br />

Left Right<br />

Min Sample Value: -32766 -32766<br />

Max Sample Value: 32765 32765<br />

Peak Amplitude: 0 dB 0 dB<br />

Possibly Clipped: 0 0<br />

DC Offset: -.037 .013<br />

Minimum RMS Power: -8.42 dB -5.46 dB<br />

Maximum RMS Power: -5.15 dB -2.08 dB<br />

Average RMS Power: -6.41 dB -4.15 dB<br />

Total RMS Power: -6.38 dB -4.06 dB<br />

Actual Bit Depth: 16 Bits 16 Bits<br />

<br />

Using RMS Window of 10000 ms<br />

<br />

http://i238.photobucket.com/albums/ff132/alexb2k/HA/bb39e1a5.png<br />

<br />

<br />

96 kHz 24-bit, Track: 09 The Four of Us are Dying, passage: 3 min 50 s - 4 min 00 s<br />

<br />

Left Right<br />

Min Sample Value: -32768 -32768<br />

Max Sample Value: 32768 32768<br />

Peak Amplitude: .01 dB 0 dB<br />

Possibly Clipped: 75 439<br />

DC Offset: -.099 .038<br />

Minimum RMS Power: -7.64 dB -4.06 dB<br />

Maximum RMS Power: -3.39 dB -.61 dB<br />

Average RMS Power: -4.83 dB -2.76 dB<br />

Total RMS Power: -4.8 dB -2.7 dB<br />

Actual Bit Depth: 24 Bits 24 Bits<br />

<br />

Using RMS Window of 10000 ms<br />

<br />

http://i238.photobucket.com/albums/ff132/alexb2k/HA/8a03bf6c.png<br />

[The Computer Audiophile]

Thanks for the information Poo! This is really frustrating. It would be so nice to hear an explanation as to why NIN did this.

[poop]

Ignorance? I don't mean that in a disrespectful way - but there seems to be so much confusion about digital quality and how best to achieve it. I plan to have a more in-depth listen over the coming days to compare the quality difference my ears pick up - but not a good sign that my initial listen didn't present any obvious improvements...

[markr]

I downloaded the CD quality FLAC and the 24/96 WAV versions, and I'm glad they've given me 'The Slip'. I am looking forward to comparing the two versions seriously this weekend. I have to say that I like the sound of both versions so far. Poo's waveform uploads have caused me to think somewhat in-depth about this a bit ahead of my critical listening excursion. Here is what I expect: ...<br />

<br />

NIN isn't exactly a chamber orchestra or a jazz quintet. Recording NIN material isn't necessarily an exercise in capturing 'traditiona'l musical dynamics played by traditional sorts of instruments - In fact, it is quite the opposite for a large part of their material. They do include a wide variety of sound sources in their material, though - an intriguing mix.... IMO, recording NIN (as with a lot of electronic music) is likely much more about 'creating a sound' rather than capturing an existing one. With NIN the sound is often, but not always, highly compressed/limited/clipped, etc - on purpose. .... So I wouldn't be too excited about the lack of visual 'quiet spots' in the waveforms of the music - it (as always) should be more about how it sounds.<br />

<br />

I've been wrestling with the the question of what the optimal recording resolutions for various of types of music are for some time now without a real solid opinion other than (in a very general way): if it is mosty non-electronic/acoustic music then higher resolutions seem to be best for accuracy; the more the music leans towards electronic sculpted/shaped sound the less the resolution it seems is required for accurate sounding recordings. It ain't all science. It's hefty doses of art too.<br />

<br />

I am also curious about knowing exactly what equipment was being used during the recording of this album. From the sound of it so far, it must have been quite a cool mix of 'toys'....<br />

<br />

markr

[The Computer Audiophile]

Hey Mark - I was thinking the same thing. NIN is all about a certain sound as the final product and whatever it takes to get there is how they'll do it. I kind of eluded to this by saying I wish I new why or had a explanation for what they did.<br />

<br />

This type of music is just different from the standard "audiophile" music that attempts to recreate a performance of artists and their instruments such as Miles Davis and his Trumpet.<br />

<br />

Don't get me wrong though, I find NIN much more enjoyable than a non-clipped / compressed perfect recording of music I don't like.<br />

[markr]

... on this Chris. I'm loving this one. I got a few minutes to get more 'near' to the music here via my iPod (16/44.1 of course) and the AE2's at work tonight (thanks). VERY nice imaging. .......WHAT is being imaged is interesting too. I am detecting some things I did not expect. I cannot wait to get some in depth-time with the 24/96 version and say, my AKG studio cans........<br />

<br />

THEN I want to examine the master tracks in their unmixed state. <br />

<br />

Yep: FUN! Sweet downloaded uncompressed digital audio is GREAT!<br />

<br />

markr

[markr]

... but not ALL of the types I can imagine doing have been done. I do however have something to say, so let's do it!<br />

<br />

Some things occur to me right away:<br />

- the 16/44.1 files must be decompressed from lossless. I am sure that has *some* effect on my ability to accurately compare.<br />

- I cannot readily A/B compare the versions, it has to be done serially - adjusting system and DAC settings, then switching files - That will color my judgement.<br />

- My A/D/A box has progressively better analog audio specs as one raises the bit and sampling rates. More effect.<br />

<br />

Loose first-audition impressions:<br />

At 24/96, it is astounding to hear what IS the clear sound of damaged audio playback hardware coming from my system with absolute clarity (no actual damage occurs either... careful with that volume control ...). Even more amazing is that while that is going on, there are fine and almost transparent instrument sequences and voices creeping around on top of and streaming through and under the 'damaged boom box' sound layered across the base of the soundfield. This is all placed in spaces of various dimension from almost infinite, to the sound of my inner voice murmuring to me from behind my hand covering its mouth. This is masterfully done. I love the recording techniques used here. The material is supremely interesting to me as well. These guys work very hard at this. I could actually disintegrate flesh and physically damage mine and my neighbor's house with this material if I aimed my Klipsch LSI's properly and set my pre's and mains to the "kill" settings.......<br />

<br />

I am not so impressed with the 16/44.1 sound initially however. It is decidedly muddier and constricted. *If you know my opinion about 16/44.1 digital from the 'get-go', this opinion will not surprise you, but it is honest on my part* The aforementioned distortion is there - not as clear-, but one has to strain to hear the overlaid fine and transparent parts that are woven into the sound of the thing. <br />

<br />

I also tried the 16 bit version with my system set to upsample to 24/96. While there was more overall clarity, it did not compare favorably with the total package that is the 24/96 version. I suppose a lot of folks would love it at this resolution. Especially NIN fans. Knowing what the 24/96 files sound like, I'm having problems with it.<br />

<br />

Now, there are a number of ways that I could have been sort of 'forced' to this opinion. I mentioned a few at the top. There is also the possibility of intentionally 'doctored' versions being offered to us as these downloads. I don't know what the truth would actually be though.<br />

<br />

Also, there are some discussions going on at the website that Poo provided the link to above - hydrogenaudio.org - that assume that Trent & the boys are ignorant to some measured and perceived anomalies of the waveforms provided by the downloaded material. I would submit here that the odds are that the ignorance lies elsewhere than with NIN or its organization. There are too many statements being made there to properly comment upon them here. <br />

<br />

Suffice it to say that when one creates music, it comes - and should come - from the source that sounds best for the intended purpose of the end product while one is there in the studio, or wherever, creating it. That doesn't always lead to the best possible track to drop on the master mix. Instruments operating in the analog domain present their problems to record properly, and the digital instrument resolution of today is not the same as the digital instrument resolution of yesterday - but some of yesterday's instruments play and actually SOUND better - or at least have 'the sound' that you want to use. NIN uses them all. This, to me, explains the perceived anomalies being discussed online elsewhere. In order to get this varied performance material on the finished product you have to 'improvise adapt and overcome.' They have done that. Very nicely. It appears to be most clearly heard on the 24/96 version IMO. It could have been done as well at 24/88.2 or 24/192. It just happens to be at 24/96 for the purposes of this distribution, or because that was the highest resolution of some of the instruments they used to perform with. I'm really looking forward to reading what Chris and Poo think about the 24/96 files vs. the CD quality ones - if they choose to do so.<br />

<br />

I've got to go do some more listening now: I'm 'cleansing my pallate' with some Stravinsky Firebird and Petrouska performed by Stokowski conducting the Berlin Philharmonic. Thanks to the 24/96 downloads made possible by High Definition Tape Transfers. I think I'll line up some Linn audio stuff too.<br />

<br />

markr<br />

<br />

<br />

[The Computer Audiophile]

Very cool Mark. Thank a lot for laying out all that detail and your honest opinion. I've been listening to equipment for review purposes lately and that is best done with music one is very familiar with. Thus I haven't been able to listen to this album for more than ten minutes at a time. Now you've really got me interested in dedicating some serious time to the album. <br />

<br />

Thanks Mark!

[markr]

.... It occurs to me now, that I should try listening to the 16/44.1 file upsampled at 24/88.2 instead of at 24/96. What could it hurt?<br />

<br />

markr

[markr]

I took my own advice and 'properly' upsampled the CD quality download of 'The Slip'. <br />

<br />

The 16/44.1 files upsampled to 24/88.2, sound *almost* as good as the 24/96 files. I believe that the 24/96 version of this album was designed to be JUST that: 24/96 ... If you want the best performance out of the CD quality files, upsample properly. This is definitive to me now: if you are going to upsample digital files, you must follow a rule of "+1.5 X the bitrate and 2X the sampling rate". Do not stray from this.<br />

<br />

This will cause many designers of DAC's to have a headache, but if they want to compete, they need to follow the rules. My A/D/A box is fully compliant, down to adjusting the sample rate at the 'cent' level. I am REAL happy with that.<br />

<br />

I AM LOVING THIS.<br />

<br />

markr

[poop]

Amazing to see, NIN have updated their .WAV and FLAC versions of the release - this is straight from their site:<br />

<br />

"** UPDATE<br />

Some friends at hydrogenaudio pointed out an error with the high resolution WAVE files we didn't catch. The corrected files are now posted along with high resolution FLAC files as well. If you're interested in these, simply re-download the new files.<br />

Sorry for the hassle!"<br />

<br />

Incredible that our discussion at Hydrogen Audio and subsequent contact of Nine Inch Nails has been taken seriously and dealt with accordingly! Imagine how this might have gone if we were dealing with a majior label...<br />

<br />

This is awesome!

[The Computer Audiophile]

This may be the coolest thing I've heard in a while.<br />

<br />

Thanks Poo!

[poop]

I'm blown away TBH...<br />

<br />

Thought I'd add this also which I found interesting, straight from Trent himself in his response to us on Hydrogenaudio:<br />

<br />

"BTW, the record was recorded at 24/96 using a Lavry AD122-96MKIII, Antelope's Isochrone OCX clock and mixed in analog through the SSL AWS 900+. We mixed back into Pro Tools through the Lavry as well as a separate rig running at 24/192 using Apogee A/D. The mixes we chose varied song to song based on what sounded best to us."

[markr]

... how much better could it sound? I just learned something about you & the gang over at hydrogen: my apologies, and Kudos to you all. Thanks for the update and detail too. I'm pretty sure that a major label would not have even acknowledged your queries on this issue.<br />

<br />

Now gotta go and download it again!<br />

<br />

markr

[markr]

... I've got to say that this sounds s-o-o-o-o-o much better now. The 16/44.1 files are much more precise, especially played back with the DAC set to 24/88.2! = vastly improved iPod quality..... The 24/96 audio is just flawless now. Great. Just Great. ..... and it was free both times!<br />

<br />

markr

[poop]

I thought the 16/44 files were the same (and left unchanged) and it was only the higher res encodings that had the problem... So long as you can hear an improvement I guess...

[markr]

... but I do hear a difference at in the new 16/44.1 files from the first ones I downloaded - I can no longer 'A/B' compare them though, because I just deleted the originals of both resolutions and then downloaded the new files to the existing folders on my hard drive when you posted here about the update..... <br />

<br />

[got tinnitus]

OK, this is all a bit late but I only just got the chance to download the 24/96 offering.<br />

<br />

What can I say; the 24/96 file/s are way better than the 16/44.1 'version'.<br />

24/96 is more, um, analog like, it's crunchier sounding, deeper depth, higher height, wider soundstage. <br />

<br />

Like, 24/96 rulz dood.