K-Meters are now "Dynamic Range Meters"?

[herodotus]

I have to say, I am not understanding the controversy here. It's not like someone is going to be standing over Rick Rubin or Bob Rock telling them to mix differently. It's just a reference number.

Personally, I think that heavy rock sounds fine if it has a 'DR' number of 9 or 10. Nirvana's In Utero generally hangs around in that range.

QOTSA's Songs for the Deaf hangs around 4 and it sounds like shit.

The fact that the application's authors like music with a range of 12 or 14 doesn't really mean that the meter has to be used to make music in that range.

[MitchK1989]

This metering system needs to be compensated for bass or we're just plain not allowed to have pumping bass without the rest of our track being turned down to nothing... Bass frequencies are not perceived as loudly as other frequencies, and this system does NOTHING to account for that.

EDIT: Also to expect every studio in the world to adopt a windows only plugin for metering is just hilarious, considering in studios I believe there is about a 50/50 split for windows/mac (it depends on whether you only count studios that make money, only count studios that are associated with labels, etc... The number can vary WILDLY... But the point is, nothing can ever be universally accepted in the music business software wise unless it runs on OSX AND WINDOWS BOTH)

[herodotus]

Also to expect every studio in the world to adopt a windows only plugin for metering is just hilarious, considering in studios I believe there is about a 50/50 split for windows/mac (it depends on whether you only count studios that make money, only count studios that are associated with labels, etc... The number can vary WILDLY... But the point is, nothing can ever be universally accepted in the music business software wise unless it runs on OSX AND WINDOWS BOTH)

We already work on an RTAS and AU version of the plug-in and a MAC version of the TT DR Offline Meter and expect that for May 2009.

[Compyfox]

guys, sorry... aes17 for k metering or not?

As far as I know, AES17 is a different standard (I think they were old weighted VU meters, like the PSP Vintage Meter).

Then again, Mr Katz can answer that better than I can
http://www.gearslutz.com/board/masterin ... 7-not.html

CDs are not the distribution system of the future, so a label on artwork is an idea from the past, most mp3s don't come with artwork.

According to the developers of the system (whomI have contact with), there is a lot planned for the future, including digital media (this is why the CEO and Inventor of the Fraunhofer Institute is on board).

Is this system going to be applied on a song by song basis, or will one DR number be applied to an entire CD?

The entire CD. This is why there is a standalone tool that reads out the whole CD/batch (unfortunately 44/16 as of this moment). According to that final readout, the CD will then be tagged accordingly.

In terms of my last album, it will look something like this (mastered in K-12/AZ+4):

Code: Select all

------------------------------------------------------------------------------------
 Analyzed folder: X:\\macera\Mitschnitt\
------------------------------------------------------------------------------------
 DR   	   Peak     	   RMS      	 Filename
------------------------------------------------------------------------------------

 DR9	  -0.00 dB	 -11.74 dB	 macera - Keiner dieser Tage.mp3
 DR8	  -0.00 dB	 -10.55 dB	 macera - Where the Streets have no Name.mp3
 DR10	  -0.02 dB	 -12.54 dB	 03-macera _ Didi.mp3
 DR9	  -0.00 dB	 -11.34 dB	 04-macera _ Bir Derdim.mp3
 DR9	  -0.00 dB	 -11.74 dB	 05-macera _ Weswegen.mp3
 DR8	  -0.00 dB	 -10.10 dB	 macera - Vergessen.mp3
 DR10	  -0.00 dB	 -12.81 dB	 07-macera _ In the Air Tonight.mp3
 DR10	  -0.02 dB	 -13.26 dB	 08-macera _ Aicha.mp3

------------------------------------------------------------------------------------

 Number of files:    8
 Official DR value:  DR9 

====================================================================================

If applied on a song by song basis, then the artistic decisions about how one song on a CD will flow into the next will now be out of the artist's hands.

Agreed.

As someone pointed out earlier, this will break "compatibility" with current and older CD releases, which will be much louder than the DR releases. Listening to a mix of the two might be a surprising and painful headphone experience.

This is where Auto-Gain comes into play and where the DR number plays an important role. But we already discussed the sideeffects on that one (and I wouldn't support that idea either). Or, what I'm still against: an additional volume adjustment before the material is pressed onto the medium.

Still a hot topic at the moment. So I'd sit and wait that one out.

This metering system needs to be compensated for bass or we're just plain not allowed to have pumping bass without the rest of our track being turned down to nothing... Bass frequencies are not perceived as loudly as other frequencies, and this system does NOTHING to account for that.

I have to agree here, and this is why I'm actually working on a larger paper for the developers, with recommendations regarding that issue.

One of them being focusing on the K-System and saying that this particular system actually "sets" the absolute loudness limit, while the DR-System is actually just an indication of what this material is actually in "dynamic wise".

If I look at the following quotes:

Personally, I think that heavy rock sounds fine if it has a 'DR' number of 9 or 10. Nirvana's In Utero generally hangs around in that range.

QOTSA's Songs for the Deaf hangs around 4 and it sounds like shit.

What is the proper dynamic range for an album like Metal Machine Music, for instance?

I'd say we create 3 dynamic ranges, like the K-System does, too. This is still a thing to think through, but last night I was pinning down some quick ideas for the time being. My ideas so far were, while taking the K-System into consideration:

Code: Select all

For this sytem to work, the K-System declares the maximum loudness (K-12/AZ+4 for Broadcast or as absolute upper limit in the first 2 years of implementation, K-14 to K-14/AZ+4 for most CD productions, K-20 for data transfer/pre-mastering), while the DR-System is just an additional indication for the program material. There is no need for a volume adjustment if both systems work hand-in-hand, since most DVDs with auto-gain are, to my impression, usually not louder than K-12.


K-12 with used up amber zone (K-12 / AZ+4 => -8dB RMS):
---
Absolute maximum for radio broadcasting.
Allowed DR Ranges: DR12 to DR8, but no higher than DR8


K-12 with half used up amber zone (K-12 / AZ+2 => -10dB RMS)
which resembles K-14 with used up amber zone (K-14 / AZ+4 => -10dB RMS)
ranging down to K-14 without used up amber zone (K-14 / AZ+0 => -14dB RMS):
---
Suitable for most CD productions including Pop/Rock/Electronic while still having the best of both worlds: hot material (if needed: K-12/AZ+2) while still maintaining the full dynamic of a song
Allowed DR Ranges: DR14 to DR8, but no higher than DR8 - recommended: DR12 to DR8


K-20 with half to fully used up amber zone (K-20 / AZü4 => -16dB RMS)
---
Suitable for Orchestra/Jazz, but mainly used for Production/Pre-Postproduction/Data transfer/cinema
Allowed DR Ranges: DR14 to DR10, for data transfer/premastering no higher than DR12


Program Material (the loudness standard is set by the K-System, no alterings needed):
DR-Range Rock: DR12 to DR8 (DR12 for classic rock, DR10 to DR8 max for higain rock, best if DR9)

DR-Range Pop: DR12 to DR8 (depending on the arrangement)

DR-Range Electronic: DR12 to DR8 (best if no higher than DR9 due to the bass intensive material)

DR-Range Classic/Jazz: DR14 to DR10

DR-Range Post-Production/Pre-Mastering: DR20 to DR-12 (best if not higher than DR12, depends on the program material - a dynamic "headroom" of about 4dB is advised -> no heavy compression on the master bus, for premastering the stereo render shall not be louder than -2dBFS)

This is by no means a standard or decided by the developers (Pleasurize Music Foundation) - this is still a rough draft from me (Studio Compyfox) as constructive critism/feedback for a possible future implementation.

I'm still open for ideas however. But I tried this myself so far, and I have to agree with MitchK1989, herodotus and Cordelia, that this system can basically work as indicator while producing so far only, and not as final standard. It needs a back-end, and I think the K-System is working best for that purpose.

So there might actually be 2 Logos: K-<number>/DR-<number>
If you combine this with color codes (K-12 in orange, K-14 in DarkBlue, K-20 in green / DR-System in LightBlue/White/Black), you have some simple indicators what the CD has as song material. So you only need 3 volume knob settings actually, and no autogain system of some sort.

As a guideline, as an educational tool, I think this system is OK, if flawed. As a mandate...

Good that the developers are still open for suggestions.

[AK--47]

*bump*

New Version out: 1.4a
check out: "Bobby McFerrin - Don't Worry, Be Happy" loud as hell and DR16

[Compyfox]

But barely any bass if I can remember, which eats up a lot of the dynamics.

[P.T.]

A system that demands a certain minimum dynamic range for music is out of the question to me.

Dynamic range is an artistic decision.

I don't like the current smashed to the point of distortion mastering, but I am completely against mandated dynamic range standards.

[kylen]

I've been thinking a lot about dynamics lately thanks to this thread and would probably appreciate some type of rating for how smashed or not-smashed commercial recordings are per some kind of dynamics standard (maybe this one). Right now I just use K-scale or dbm metering tools like IXL and SPAN and ears. I'm going to demo these dynamic range meters soon...

I didn't think this was some type of constraint though, I'd agree with P.T. on that point.

[eduardo_b]

A system that demands a certain minimum dynamic range for music is out of the question to me.

Dynamic range is an artistic decision.

I don't like the current smashed to the point of distortion mastering, but I am completely against mandated dynamic range standards.

I understand your point about artistic decisions, but I don't think this qualifies. It's just like the RIAA equalization curve. It serves a purpose beyond where artistic decisions get made. It's about audio quality standards, within which artistic decisions can be made while giving listeners the opportunity and right to hear the music without ear fatigue and half the dynamic range.

[Compyfox]

While I do agree, I listened to a lot of material recently with the DR-meters on. It is a fact that most DR14 to DR-11 stuff is either under-EQed, fairly clean stuff (not bass intensive hiphop/dance/hi gain rock) or recordings from the pre 80ies.

I do agree that there has something to be done about overcompression, but this is why the K-System is for and is, in my humble opinion, the key element to the DR-Metering.

Of course we can use the RIAA vinyl cut stadard. This way we have full dynamic, but no bass at all, which is being dialed in again by the amplifier. Going back in time is one thing, having dynamic in a track again another. But not having thought this through and saying that "DR8 tracks need to be adjusted back down to DR14) is just not helping.

Meters respond more to low frequency material, and this is a major flaw from weighted systems (VU meters, AES17, this DR-meter, the K-System), but the K-System actually offered the best of both worlds.


I'm not throwing any more sand into the gear as there already is, but I really do hope that some stuff will still be thought through before anything is finally set. And for this, I'm looking forward to meet the initiators of this system at the next music fair (in this case, Musikmesse 2009).

There's still some time till Summer 2010 after all. And it's good that the developers are not unresponsive to constructive critism.

[eduardo_b]

While I do agree, I listened to a lot of material recently with the DR-meters on. It is a fact that most DR14 to DR-11 stuff is either under-EQed, fairly clean stuff (not bass intensive hiphop/dance/hi gain rock) or recordings from the pre 80ies.

I can't tell if you mean underEQed in a general sense or a negative sense. As for the pre 80s, there are any number of engineers who believe the golden age of audio engineering in terms of quality was the 70s. I'm sure that is open to debate in the larger sense, but I have fond memories of many great albums that sounded wonderful on a quality audio system -- in old school terms, high fidelity.

I do agree that there has something to be done about overcompression, but this is why the K-System is for and is, in my humble opinion, the key element to the DR-Metering.

The K-System would be fine if it was being used as a standard within the industry. But it's not. Something more needs to be done, and perhaps the dynamic range approach will raise the issue in a way that loudness doesn't, can't or won't.

Of course we can use the RIAA vinyl cut stadard. This way we have full dynamic, but no bass at all, which is being dialed in again by the amplifier. Going back in time is one thing, having dynamic in a track again another. But not having thought this through and saying that "DR8 tracks need to be adjusted back down to DR14) is just not helping.

I was only referring to the RIAA curve as an example of having quality standards within which artistic creativity has ample room. I can see it being possible to reissue CDs from the last decade plus as remastered, but in this case having the dynamic range restored to a greater degree than in crushed releases but perhaps not as great as in preceding decades.

[Compyfox]

I can't tell if you mean underEQed in a general sense or a negative sense. As for the pre 80s, there are any number of engineers who believe the golden age of audio engineering in terms of quality was the 70s. I'm sure that is open to debate in the larger sense, but I have fond memories of many great albums that sounded wonderful on a quality audio system -- in old school terms, high fidelity.

Actually I did mean this as negative sense. Stuff that just needs a bass (since it's electronic) or rock material, where distorted guitars barely have any dynamic, would suffer from this particular system (in my opinion) if you're forced to use 14dB as dynamic range. Granted you have enough headroom, no doubt about it, but the meter responds to the loudest parts, and if they're bass intensive or due to the guitars less dynamic, it declares the whole dynamic range for the production. This doesn't work.

While I also like recordings from the 70ies (Hendrix, Pink Floyd), the thing is, the old sound is a history thing. Nowadays you've got to stick with the flow, else you don't sell anything. And in this case, it is a fact, that the bass is mixed moderately to not being heard at all. The whole production lacks bottom end.

And this is my major problem with this system - I think the negative sideeffect is going back to the sound of the 70ies. And for a lot of productions, that isn't working (which is why I work on a paper with several sub-standards at the moment).

The K-System would be fine if it was being used as a standard within the industry. But it's not. Something more needs to be done, and perhaps the dynamic range approach will raise the issue in a way that loudness doesn't, can't or won't.

I think the K-System can be used as backbone for the DR-Meters, or the DR-Meters as backbone for the K-System. Combine both, and you have a wonderful meter while producing and postproduction, and the K-System meter while metering, leveling in and while mastering (standard maxima, etc).

This is why I also mentioned earlier that I'd suggest a double-logo system. The K-System tells you how loud you can turn up your amp, while the DR-Meters give you a "general" overview of how dynamic the material is. And in this case: Dance tends to be in DR8, same goes for heavy metal, while Classic can of course be DR14 down to DR8 (Gregson-William's Soundtracks from Sindbad and MGS: Snake Eater were in DR9 btw).

Both systems thrive for the same goal, and both systems need each other. It's only that the Pleasurize Music Foundation actually makes a move towards the industry, while Bob Katz was hesitating in terms of Logos and the likes. But I encourage both fractions to work together - since this is what might actually work IMO.

I was only referring to the RIAA curve as an example of having quality standards within which artistic creativity has ample room. I can see it being possible to reissue CDs from the last decade plus as remastered, but in this case having the dynamic range restored to a greater degree than in crushed releases but perhaps not as great as in preceding decades.

Well the RIAA Standard has the disadvantage with the bass dip. And CDs have the disadvantage of sounding too clean, but they have a better frequency response (which tool like 10 years to be handled well), but the engineers were all like "Let's use the very final bit of this thing".

I blame all sides of the table that it went out of hand. But simply dropping everything that was discovered in the last years and going back to RIAA cuts, is just a bad decition in my opinion. Again, this is just an example, but it's going into that direction.

I agree with a clean restart, but then I'd not say "a track shouldn't be less dynamic than 12dB (from overall loudness at the loudest parts to highest peak)", I'd rather say we were presented three loudness standards years ago, why not use them, and another 'standard' in progress as additional indicator".

I mean, it works for DVDs and BlueRay, why doesn't it work for audio material, too? It's less effort than redoing CD/MP3 player hardware and implementing an auto-gain feature that is calibrated to the DR-Metering system.


Then again, maybe I totally interpret it wrong. But so far, I wasn't proven wrong either.

[eduardo_b]

Actually I did mean this as negative sense. Stuff that just needs a bass (since it's electronic) or rock material, where distorted guitars barely have any dynamic, would suffer from this particular system (in my opinion) if you're forced to use 14dB as dynamic range. Granted you have enough headroom, no doubt about it, but the meter responds to the loudest parts, and if they're bass intensive or due to the guitars less dynamic, it declares the whole dynamic range for the production. This doesn't work.

The overall issue seems to be that there were stopping points along the way that would have preserved the appropriate dynamic range and retained reasonable loudness, and each of these points was ignored because all of the emphasis was on the obvious measuring scale of loudness. That clearly hasn't worked because there has been no incentive among many in the industry to pay attention to the overall qualities of production and post-production beyond loudness.

The problem is that personal opinion isn't a scale that's reliable, so another more rigorous method is needed. Some relationship between loudness and dynamic range that can be measured is the only viable way of avoiding the excesses. Overall, loudness will return to levels of perhaps the late 90s or earlier in that decade using such a system.

What won't work is allowing current loudness levels to be maintained as the standard. One possible way of dealing with this, and has perhaps already been raised, would be to allow a logo on those commercial releases that meet the dynamic range/loudness standards that are finally set. This would have the effect of giving an educated consumer (educated via campaigns regarding this) the opportunity to not buy CDs or music that doesn't meet the standards necessary to use the logo. Those who push loudness too far will be punished by those who won't buy music that hasn't met the standards for displaying a logo. Might work...actually will work if consumers and enough industry types sign on to the concept and use their influence and money to affect opinion.

[Compyfox]

Yeah but for this alone, you need "more" than just a dynamic stanard.

Like I said over and over... Electronic/Rock/Pop... this all has less dynamic, even in K-14 (which is -14dB RMS) than let's say a classic track. If the production is in the K-System already, which conserves the dynamics of a track, does it really need to be altered afterwards cause the dynamics are at the standards boundaries?

This is what I'm talking about. Alone it doesn't work - both standards (K-System and DR-System) need a backbone. And I really think they complement each other.

If a track is mixed/mastered in the K-System already, there is no further need to post-alterings (reducing the volume) if a track is in DR9 or something, because the style of the song is just like that.

It's working if used both systems. But using the DR-Metering alone, it doesn't stand a chance, neither does the K-System as long as nobody is actually pushing it (like the DR-Meters are pushed at the moment).

This is the very reason why I said:
Combine the systems, create a dual-logo like "K-14/DR-10" to say "turn up your amp to this'n'that value, and be prepared to hear a hi-gain rock production" or like "K-12/DR-8" would resemble "this is a hiphop CD, still in a suitable loudness but has a more limited dynamic"


At the moment, the DR-Metering looks to me like a brandmark over half the cover, like recently done over here with videogames, where people say "OMG! DR-8 is like FSK-18 - you're NOT allowed to play that". It's just aiming at the wrong direction IMO.

[eduardo_b]

Combine the systems, create a dual-logo like "K-14/DR-10" to say "turn up your amp to this'n'that value, and be prepared to hear a hi-gain rock production" or like "K-12/DR-8" would resemble "this is a hiphop CD, still in a suitable loudness but has a more limited dynamic"

I don't see that as being practical or particularly useful if it's tied to amp settings. The presence of the logo says this is an appropriately produced album for the type of music it is. I agree on the two systems working together, but the music-consuming public isn't going to deal with specific settings on car or portable players, or even home setups. It will be quite enough an accomplishment to get sufficient numbers of those within and outside the industry to agree to having such standards by type of music -- we assume they categories can be also agreed upon -- and having this information displayed in a consumer-friendly way -- some symbol/logo that's a stamp of approval on audio quality. This would have the potential to eliminate the worst aspects of the loudness wars, bring back appropriate dynamic ranges and improve the listening experience.