Is the DR database really accurate for vinyl?

Right... the people I mentioned (Sax, Ludwig, Hull, et al) strongly recommend to their clients a separate mastering for vinyl so that they can make more money. Or perhaps they just want their clients to have the best sounding vinyl LP possible. I guess we can all decide for ourselves. BTW, I thought it was considered bad taste for professionals to criticize other professionals in a public place like this.

Just for the record, I quoted you properly. Re-read post #209.

When it comes to music like that, acoustic instruments, vocals, with minimal accompaniment, etc, I prefer to hear a very natural, lifelike sound, as if the performers are in my living room. I didn't sense that from this recording, but as I mentioned before, I'm listening to sound over the web, and I wouldn't make critical judgements based on that.

There are so many... I've got hundreds of jazz and classical, and even pop LPs that sound as if there was no compression used in mastering - but of course I wouldn't know for sure unless I was there or the engineer made a statement about it... but it's certainly not uncommon for records to have a good 15db or more of dynamic range. Lately I've been enjoying some of the Music Maters 45RPM Blue Note reissues. And I am not categorically against compression either. We all like what we like... no one should feel a need to defend it.

We're all entitled to our opinions, all we can do is try to state them as honestly and politely as possible. And it's fair game on a forum like this to debate how "sound" any mastering advice is.

 

Uh oh can I edit my post.....

 

Hm, that was supposed to be a reply to the post where you said

I must have clicked "reply" on the wrong post, sorry.

You're right about "Death Magnetic", as I have every reason to know:

http://mastering-media.blogspot.co.uk/2008/09/metallica-death-magnetic-sounds-better.html

 

This supposedly is NOT compressed at all, although I think maybe the instruments might be before going TO the tape somewhat.

Very dynamic, but not according to DR.....lol



Turn your volume way up and play at 720P!

 

I haven't read the entire thread, so i apologize if this had been covered. There is a possibility that if the record had successfully been cut with square wave information at peaks (specifically high frequency), that the stylus on playback does not react quickly enough to re-create the clipping accurately, resulting in a waveform without flat peaks. This could add transient information, changing the DR result as well.

 

I'm confused. You quoted me saying I was sure that wasn't the case for the people you mention. I wasn't being sarcastic, I meant it. However there are many other vinyl engineers in the world...

I agree, and I haven't criticised anyone. Masterdisc have been responsible for some of the most heavily crushed, "loudness war" releases of recent years - that's a fact. They also heavily promoted the bogus concept of "adaptive mastering" for lossy compression formats like AAC, misleadingly implying that was what "Mastered for iTunes" is all about (it isn't) and doing so on an intensely over-compressed release (RHCP "I'm With You") - also a fact. So for them to be evangelising dynamic range is ironic, as I said.

However, Scott Hull was not responsible for any of the releases I'm critical of, and of course there are often other pressures on engineers beyond their personal taste and judgement. Again, I said its ironic, that's all.

Fair enough - and, I agree. In my opinion the compression I used enhanced those properties, but you'll have to take my word for it. Any master can only ever build on what was present in the original recording, of course.

No, and that's my point - few people can detect compression without comparison to the original source, if its well done. You may well be one of those people, of course !

You mean on CD, presumably ? Yes, and I typically master jazz with a good 3dB or more PLR, if I'm allowed.

Of course ! That's what we're doing, right ? It's important to me that people know the position I take is not black-and-white. Not all compression and limiting is bad, and managing dynamics is a crucial part of recording, mixing and mastering music - and that's not at all in conflict with the idea of opposing the loudness wars.

Ian

 

Just to be clear:

I used what some people call a "brick wall" limiter on the song in the video. This does NOT mean I "admitted" to "brickwalling" it.

 

All music has compression. Even a lot of jazz stuff, but often it is inherent to the recording session. Compressors are almost always used on voices and instruments, as a way of controlling, the loudness, so it does not vary overly much and get annoying.

You want to hear no compression, go to a High school talent show and listen to a 16 year old girl sing on stage amplified. 45-60 db of dynamic range sounds cool in theory, until it is someones voice, and it is wavering all over the place volume wise, from mildly audible, to horrendously loud and laughable. ( real world experience)

 

I think that's over-stating it. The TT meter is a useful way to broadly assess how squashed the loudest moments of a digital source are - ie. how much peak-to-loudness ratio there is. Since this is the factor that most clearly tracks how "damaged" a recording is by excessive compression, limiting or clipping, I think it certainly has its uses.

However I agree, the name of the meter can be misleading - it doesn't actually measure "dynamic range" in the strict technical sense.

 

My apologies then. It sounded sarcastic to me. My fault.

I could sit here all day and give examples to make the point - for example I've heard plenty of re-masters, that did not use compression, compared with the originals that did, and I always prefer the ones with no compression used in mastering. See for example some of the recent remasters done by Steve Hoffman on CD for Audio Fidelity (to name one of many examples). Also, the reverse is of course sometimes true - when a remaster uses lots of compression compared to an earlier mastering -- see for example Barry Diament's masterings of Bob Marley for Cd in the 80s, compared with the remasters done later. I prefer the earlier masterings, which used no compression. And on, and on we could go...
Incidentally, I have done recording, mixing, and mastering myself, since the mid 90s (nothing anyone here has ever heard) So I understand the technicalities. I was also a fan of using peak limiting and compression in mastering, but my opinion has changed.

I meant for LP, but of course the same holds true for CDs. Not sure why you would think it doesn't (if you do think that).

 

Spoken like a gentleman - no problem.

Well, remasters are something of a special case - I'm much more sparing in this situation, because often you're working from an old production master which has already had some compression - but I take your point.

Just to be clear, I'm not a "fan", either. No-one is happier than me to get a master where only a little EQ and some level tweaks are needed...

You quoted DR values, which implies you've used the TT meter to measure them (otherwise how would you know ?). But this thread was started by my video which shows that those readings aren't reliable. But of course there are recordings that measure DR16, especially in the jazz genre, so I'm not going to argue the point on that.

Ian

 

The TT meter may not tell us much but it does tell us the difference between average RMS and peak RMS and it's not lying when it says the range is 4dB bigger with the analogue waveform measured by Ian Shepherd and the digital waveform - you can see that by looking at the waveform in the video. I'm intrigued to know how this could be - in fact I'm so intrigued that I've ordered a copy of the vinyl album, which comes complete with lossless digital download in CD resolution, so that I can do my own tests on it and report back here. It will be a long wait, though, because it has to come to me in the UK from the US and will probably get held up in customs.

Now this is what I like - theories that could explain the mystery. I don't think this one holds up though because I already have the digital download and from the brief look I had, there's no clipping and no flat tops. Ian Shepherd would have to confirm of course but that's what I see. Good suggestion, though.

 

By looking at meters (VU, PPM, RMS, peak, etc.) Of course you can't get a precise number that way, but no hardware or software can generate a precise numerical value for dynamic range for a song... there's too many ways to measure it over time.

 

I find it funny and ironic that some people endlessly tout the technical superiority of digital vs. analogue media - things like signal-to-noise ratio, theoretical dynamic range - as explanations for why those who prefer the sound of vinyl are just liking "distortions" and "artifical warmth". Yet when something like DR comes along and appears to support at least something tangible to explain why a significant portion of audiophiles actually prefer vinyl, it is dismissed as a faulty measurement.

Personally, I have found some correlation between measured DR and sound quality. When something sounds squashed and I run it through the TT meter, I do find that its DR usually measures quite low. I readily recognize that many more things go into the ultimate sound quality but the DR number has rarely surprised me. Certainly it is way more useful than THD of an amplifier.

I heard it described in another thread, and it kinda makes a bit of sense when seeing what the number is actually measuring, that the DR number is more a measure of microdynamics than macrodynamics. So things like subtle dynamic shifts, natural attack and decay, instrumental timbres...will be more affected by what this number is measuring. Which makes sense to me. To my ears, it is precisely things like this which sound "squashed" in low DR recordings rather than our traditional definition of "dynamics" which is really macrodynamics. And it is precisely these sorts of attributes which usually sound better on vinyl vs. digital to my ears.

 

TT DR does not measure microdynamics, IMO. It measures logarithmic peak-to-RMS ratio only in the loudest portions of the analyzed material which are likely to contain the highest peaks and are most susceptible to effects of compression/limiting applied (if any). The higher the ratio, the better the TT DR value. 5 times ratio (i.e. 14dB) is regarded good; higher ratio - great. Less then 2.5 times ratio (i.e. 8dB) is regarded bad. Anything between 2.5 times & 5 times (8...14dB) is transitional. Good TT DR value is really about maintaining healthy (5+) average peak-to-RMS ratio throughout the whole recording, but specifically in its loudest parts (upper 20% of all blocks with highest RMS values); nothing more...

 

I can't see how a "separate mastering" for vinyl cannot be done. CD and vinyl are two completely different mediums and consequently separate masterings are required. And this is regardless of whether the source is an analog mastertape or a digital file.

 

Ah, OK.

 

Yes I'm aware of that. I'm trying to put meaning to these numbers in terms of what we hear as a result of them. Clearly, in my experience as well as many other people's including the demonstrator in the OP's video, it does not consistently correlate with how DYNAMIC the recording sounds in the traditional i.e. macrodynamic sense.

 

One way to help investigate what's causing the difference in DR for vinyl vs. CD would be to instrument the code that calculates the DR value so the code spits out statistics and info about each of the 300ms bits of music that it is analyzing. My suspicion is that if you looked at the waveforms and RMS and peak data for each of those little 300ms windows of music that you'd see something different in the vinyl data vs. the CD data. That difference might offer an explanation as to what is going on that causes the vinyl to measure differently than the CD.

 

My feeling is that the DR measurement is probably correct - but that the vinyl medium has by some means artificially increased the dynamic range of the source. Given that the dynamic range of a source cannot be naturally expanded, any increase that can actually be measured can only be the result of an unnatural process. In other words, the vinyl playback is altering the source and comprising accuracy, given that increased dynamics can only come about through some sort of extrapolation of the original dynamic range.

 

My feeling is that the DR measurement may or may not be correct, but it is way, WAY easier to artificially lower it than to raise it.

 

It is kind of Ironic how much importance is given to this one DR number, on a site where all measurement are usually given very little regard.

I find the number not totally useless, but only useful enough to compare some aspects of one recording to another version of it, but only as an idea of how it "might" compare.

In this instance, listening is the only real way to compare, until they totally change the DR system to something that compares to reality.

As it is now, it is just a way to look at the loudest part of a song, and only fairly useful IF the song tends to be of a constant volume, and stay at that loudness naturally to begin with.

It is not measuring Actual dynamic range OR microdynamics. In fact it is only showing if the song is fairly average in loudness or a bit more varying in loudness ONLY for the loudest part of a song.

It ignores anything below 20% of full signal. My ears do not.

 

You have a good point there and I have to agree.

But the third option, could also be, that the DR meter with vinyl on some playback, may not be able to separate frequency response errors and more importantly distortion components from the actual musical signal on the record.

Even a relatively low amount of distortion percentage on a high frequency tone, can easily create a "signal" that can be several db louder, be measured as louder, be heard as louder, but ultimately not be the actual musical sound that was pressed into the record.

I'm not saying this IS the answer, but without eliminating the distortion aspect of high frequency tones, it seems possible for sure.

 

I only give credence to measurements which correlate with my listening experience. While certainly not perfect, I have found reasonable correlation between DR numbers and what I hear.

Again, attempting to correlate with what I hear does suggest an effect on microdynamics from lower DR. And it also kind of makes sense theoretically. If the differences in volume between RMS and peak are being attenuated I would think this would be one of the most noticeable effects. Instruments and sounds will not have their usual attack and decay, it'll all sound more "splattered" onto the soundstage.

 

I find it hard to believe that if this pretty big difference in DR was due to nothing more than distortion, the end result would sound anywhere near "good" - let alone better to many people's ears.