Article on headroom?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

pstamler

Well-known member
Joined
Jan 24, 2005
Messages
1,509
Location
St. Louis, MO, USA
Hi folks:

Back in the 1970s I recall reading a most interesting article on how much headroom was required over nominal level to avoid clipping. The author reached the conclusion that 17dB would be adequate a large percentage of the time.

Regardless of whether the author was correct (that's an entirely separate thread), my question is, who wrote the article? For years I remembered, and told people, that it was Robert Cordell, then with Bell Labs. But I've recently exchanged e-mails with Mr. Cordell, and he said he remembers the paper but didn't write it. Cordell wasn't sure who did, but suggested Richard A. Greiner.

I tracked down Prof. Greiner, now retired from teaching and living in Wisconsin. He also recalled the paper, and also said he didn't write it, and doesn't remember who did.

So who did write it? I need a citation for my class. If it was in the JAES I'll be able to dig it up with an afternoon of browsing at the Washington U. library, but I'm not sure it was; it may have been in Audio, which isn't archived anywhere around here. (It was, but in the mid-70s students began stealing the magazines and they had nothing to send to the bindery at the end of the year.)

Can anyone here help? Inquiring students and faculty want to know!

Peace,
Paul
 
I have certainly been a close student of this topic since the mid-late '70s when I got my first patent on a peak/VU metering system. I recall numerous discussions about crest factor and headroom in the recording magazines and technical publications, mostly in the context of either specific recording media (and NR systems) or signal processing hardware technology (consoles and the like including digital vs analog). I even wrote one column peripherally on the subject, in a comparison between headroom and S/N performance between +4 dBu and -10dBV hardware (in the early '80s).  I also broached it indirectly in my article on console performance limits around the same time, as well as sundry other console designers wrote in their own articles. I think Buff wrote at least one decent article on the general subject in support of his level meter that he sold for a while (look in RE/P magazine).

I don't recall a single definitive or classic article, and don't personally believe there is a single answer. IMO the best treatments of the general subject for you to expose your students to, is probably Bob Katz's K-metering system as he specifically details different target or nominal headroom requirements for different media and/or genres of recording. 

K_Meters.gif

http://www.digido.com/level-practices-part-2-includes-the-k-system.html

I hope this helps. I think there is a lot of discussion out there, but Katz comes closest to a coherent definitive overview, and his is more about how to use sundry technologies available rather than touting a single technology.

JR
 
Maybe you're thinking of something published by Neuman (or Gotham Audio, the importer here), I wouldn't be surprised if Temmer wrote something on the subject but I didn't find it in a google search.

http://milbert.com/articles/tubes_vs_transistors    this paper is about a different topic but lists some typical peak outputs from microphones while I don't see a specific mention of wrt nominal.  They do mention another study done by "Fine Recording" (about mic outputs)  but don't list it in their footnotes.

Surely some discussion in one of these... http://www.microphone-data.com/library/bibliography/

JR


PS: You might want to use Greiner for his excellent (definitive?) paper on speaker wire published in the JAES. Cordell was likewise pretty well know for his work in low distortion circuit design.


 
JohnRoberts said:


http://milbert.com/articles/tubes_vs_transistors    this paper is about a different topic but lists some typical peak outputs from microphones while I don't see a specific mention of wrt nominal.  They do mention another study done by "Fine Recording" (about mic outputs)  but don't list it in their footnotes.


That's the famous (or infamous) Russell Hamm article; I have that. It's not the one.

PS: You might want to use Greiner for his excellent (definitive?) paper on speaker wire published in the JAES.

I may well do that; I'm a few weeks away from talking about speaker and wires, though.

Cordell was likewise pretty well know for his work in low distortion circuit design.

Cordell is actually doing a workshop at our AES Central Area Student Summit on campus in a couple of weeks...at the exact same time that I'm doing one at the other end of the building. Sigh. Maybe someone will tape it. His new book on power amp design is fascinating; now I need to read Self just to have both experiences in one year.

Peace,
Paul



[/quote]
 
I was thinking perhaps something written by Steve Temmer, (also from Gotham), he was IIRC big on making absolute pronouncements back in the day. While I didn't find any thing in a quick google search.

Again and IMO the Bob Katz piece is as close to definitive as there is for a discussion about headroom in the context of recording. All other discussions were wrt a given media, technology, or genre so not as comprehensive IMO. Bobcat ties them all together as much as is reasonably possible. While arguably his focus is on the less dynamic finished output product and not the raw input stems.

JR
 
As I recall at the distance of 30+ years the article was primarily aimed at electronics -- necessary headroom in processing equipment and power amps.

Peace,
Paul
 
pstamler said:
As I recall at the distance of 30+ years the article was primarily aimed at electronics -- necessary headroom in processing equipment and power amps.

Peace,
Paul

Not to turn this into a debate but this discussion has probably evolved over the decades with popular technology..  Tape and tubes were more tolerant of overload. Use of many close mics, mixed together make a difference in both directions, but mostly towards dealing with higher crest factor or peak levels in close mic'd individual stems. Finally digital media, appears at least on the surface to be less tolerant of overload so requires more attention to crest factor, or specifically peak levels.

One could even argue (I'm not) that nominal 0VU is an archaic artifact of slow average meters, and benign overload media and audio paths. The trend in digital mediums to meter peak below clipping suggests a diminishing role for nominal 0 VU.

Even a definitive answer for 30+ years ago may be a little dated. My recollection is that professional paths back then were targeting 20 dB headroom above nominal 0 VU, with some premium models offering more, or claiming they did.

Semi-pro gear running at -10dBV while a little short on S/N and typically running on narrow magnetic media, was not really a slouch for headroom because of the cooler nominal 0VU, except of course for the cheapest, single supply executions, that weren't really that much cheaper.

Note: Power amp headroom is a whole 'nuther can of worms, with the discussion centered on differences between short term and long term loudspeaker overload characteristics. In much of live sound reinforcement clipping is so routine that most SR amps include dedicated clip limiters, that momentary neck back gain in response to clipping, with fast attack and fast release time constants.

JR


 
Back
Top