Testing audiophile claims and myths
Nov 4, 2018 at 4:42 PM Post #10,306 of 14,606

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,168
It's only important to the mixers preparing the deliverables. It's irrelevant to consumers. I got confused by this when I first looked into it too.

Most of the time it should be irrelevant to consumers. The two exceptions are when consumers manually calibrate/play with the sub and in the case of SACD, where the ".1" is not +10dB in-band gain.

[1] The selected quote I used was to show my reading comprehension: which clearly states the LFE is calibrated at 95DB on the RTA.
[2] Gregorio then quotes a source saying "if RTA is not available"....and so then obfuscates the original topic.

1. Yes, unfortunately you did show your lack reading comprehension! Your quote clearly states the LFE is calibrated to 95dB on a band-limited RTA when your main speaker is calibrated to 85dB on with the same RTA. IE. When the main speaker is calibrated to about 89dBC on an SPL meter, which it should NEVER be, NOT even in a cinema (where it should be calibrated to 85dBC on an SPL meter) and certainly not in a home/consumer environment, where it should be calibrated to 76-79dBC!

2. Gregorio then quotes the publication by Dobly themselves on how to calibrate the LFE output! Which provides the rough level of the LFE output relative to the level the main speaker is calibrated to. Maybe you just don't understand that's what the +10dB is refering to? It means 10dB higher than the calibration of the main speakers (which is 85dBC for cinemas, not 89dB!), resulting in the sub/array having an output level of roughly 91dB! Why don't you "show your reading comprehension" on that?!

G
 
Nov 4, 2018 at 4:45 PM Post #10,307 of 14,606

Joe Bloggs

Member of the Trade: HiBy / EFO technologies Co
His Porta Corda walked the Green Mile
Joined
Oct 21, 2001
Posts
11,280
Likes
3,065
Location
Hong Kong and Melbourne
Ah. Sorry. I thought you didn't know that it didn't affect playback. It's nothing that non-engineers need deal with, so I don't pay attention to it. I keep my eyes peeled for info I can use.

I reckon it just might be info I can use if my "AVR" looks like this... :gs1000smile:

speaker VSTHost system4.png


What I WOULD be interested in knowing is, supposing all speakers give flat in-band response, what should the single-tone response level be for the LFE relative to the other channels given the same digital input level? Should it be 10dB higher? 5? ??

This is a much more relevant question to me because I mix down the sub with centre and recalculate sub output by crossing over my specified LF part of all channels to the sub. So the answer to the above is the number by which I should be amplifying the LFE channel before mixing it into the centre channel, correct? Does this apply to multichannel music as well?

Sooo I think we've ascertained that the answer to the above is 10dB--if at all? Would the requisite boost be already done by computer video player apps like say Netflix going out to a (virtual) sound card?
 
Nov 4, 2018 at 7:18 PM Post #10,309 of 14,606

bfreedma

The Hornet!
Joined
Feb 3, 2012
Posts
2,618
Likes
1,698
1. Wouldn't a reinforcing mode be even more damaging to integration than a null? And if you were doing the crawl without an RTA, wouldn't your ears tend to end up choosing the spot with the most obvious sounding mode?
2. Somewhat related to the above, are we overestimating the number of AVRs out there with EQ features powerful enough to take care of a reinforcing mode in sub frequencies?


1. Not sure I agree. You can EQ down a reinforcing room mode, but no amount of EQ is going to fix a null. Would much rather be EQing down than taking the risk of trying to overpower a null which could be damaging to the sub amp without any real chance of addressing the problem. Agree that doing the crawl with an RTA is optimal, but wonder how small the percentage of HT owners own one let alone know how to interpret it. No doubt, not a perfect methodology, but I’ll still take the “sub crawl” over random placement. Particularly if you lack quality measurement tools and/or the knowledge of how to interpret the output.

The above is also why running multiple subs is highly beneficial, as it’s unlikely that both will be hampered by the same room issues. A lot of what I’ve posted is based on running multiples and I probably should have mentioned that earlier for context.

2, I don’t think so. Other than a few boutique receivers, it’s been a while since I’ve seen an AVR that couldn’t competently address bass room modes. The last major manufacturer that had EQ that didn’t address bass properly was Pioneer with its older version of MCACC (did nothing below 80Hz) but that was updated some years back. Emotive (sorry Keith) also released some processors that incorrectly handled bass, but I believe their most recent model resolved the problems. Additionally, a number of better subs include their own EQ. JL Audio’s implementation has worked well for me and I’ve seen SVS’s work successfully. Typical best practice is to run the subs native EQ solution first, then the AVR to reduce the demand on the built in AVR EQ.

Edit: I mean this as a compliment - after seeing your “AVR” above, I can understand why you might have concerns about the capabilities of consumer products. Do you have any documentation about that configuration you can link to? I’d love to read about what you’ve got there!.
 
Last edited:
Nov 4, 2018 at 10:06 PM Post #10,310 of 14,606

Steve999

smooth, DARK
Joined
Jul 15, 2002
Posts
2,481
Likes
303
Thank you so much, Gregorio. This is so helpful and insightful I wouldn't even know where to begin. I did not quite get the distinction between deliberate LFE frequencies and the subwoofer just picking up the lower frequencies the mains can't handle and my subwoofer doing double-duty in that respect. I had an uneasy feeling there was something I didn't know but now it is clear. I have a 5.1 system but no 18 inch drivers in my mains. : )

Also the lowered reference calibration value for music as opposed to cinema meets with my subjective impression very well, as does the feeling that once you go below that low E
(40 hz or so) it's atmosphere and shock waves (or else a pedal note, as you say).

The insights about the different genres and the low frequency content handling and the different ways of getting that low-frequency kick is very cool. Thanks.

1. Your observation is very valid IMO. 40Hz appears to be about the point at which insensitivity to amplitude/volume, directionality and frequency combine to make anything below that essentially pointless musically. Some instruments can go lower but very rarely do and when they do, I can't think of a single example where they are not just playing a pedal note, IE. A note an octave below the bass/fundament pitch of a chord, just to provide some almost intangible depth/richness.
2. I am a movie guy but also a music guy. I started as a trained orchestral musician and then was an orchestral musician for about 7 years, then I switched sides and started composing and producing music for film and TV before gradually moving over more to the sound side of film/TV but I still have to deal with music daily and still occasionally take on music production or mastering jobs.
3. I'm not sure that would/should be of any help to you, probably more of a hindrance! Most of what I stated about calibration levels is theatrical multi-channel sound, all the +10dB in-band stuff is NOT applicable to a bass managed stereo system, only to how the LFE channel in 5.1, 7.1 or the newer formats should be output.

A clear point of confusion for some is the ".1" term and it's not surprising it's confusing because it actually has two different meanings/uses! In home systems, say a 2.1 system, the ".1" doesn't exist in the music you're playing, it's simply manufactured by your bass management system and comprises all the frequencies below a certain crossover frequency. This ".1" is then fed to a subwoofer and the end result is hopefully a balanced sound throughout the spectrum, with the subwoofer effectively filling-in the low freqs which your main stereo speakers can't reproduce accurately (or at all). This is NOT what the ".1" is for in multi-channels formats such as 5.1, 7.1, 9.1! These formats were all originally invented for cinemas, not home use. In a cinema we have a bank of subwoofers, sometimes 16 or more, each of which is a 12"-18" driver but this bank of subs are NOT there to fill in a LF hole that the screen/main speakers can't reproduce because they CAN reproduce them, abundantly and accurately. In fact, each of the main/screen speakers has at least one and often two 18" drivers of it's own. If we played say a symphony through the main/screen speakers ONLY, the result should be a perfectly balanced reproduction throughout the spectrum, all the way down to 30Hz (if there is anything much at 30Hz in the recording). And indeed, when I was only supplying the music to be mixed with the rest of the film, it was always mixed in 5.0, the main speakers providing all the bass I ever needed, even when I needed a lot! So why do we need that additional array of subs? Only for effect, to move large amounts of air in the low freqs, to give the audience an actual physical impact sensation, essentially mini-shockwaves. In other words, this sub array exists ONLY for these Low Frequency Effects, hence why the audio channel supplying these subs is called the LFE channel. In a home environment exceedingly few have the type of full range main speakers used in cinemas (with subwoofer/s effectively built into the speakers), consumers use bass managed systems and this is where the situation becomes confusing, because their subwoofer is essentially doing double duty, playing both the ".1" which covers all the freqs their main speakers can't, plus the theatrical ".1" which is the LFE channel. And just to make sure it's really confusing, the theatrical ".1" needs to be output at a much higher level than the bass management ".1"! One last point I'm not sure if you picked-up on, the 85dBSPL = -20dBFS calibration level only applies to cinemas, in the home it should be about 76-79dBSPL = -20dBFS (depending on the size of room) and this applies specifically to TV/Film, music masters are NOT calibrated to the film or TV specs and almost all music, with the possible exception of some classical music, will be too loud, even at the 76-79dB = -20dB level! For most modern popular music genres, probably somewhere nearer 64dB equalling -20dBFS would be more appropriate or as high as about 70dBSPL if you like your music loud.

4. No, you're not wrong but it is rather genre specific. All popular genres going back to the 70's have a highly processed kick sound but with modern genres it's not only processed but often more manufactured than just processed. With modern R&B for example, a defining feature is a highly processed and compressed kick that's mixed with some very LF tones (typically around 40Hz but can be lower) giving it that longer duration boom sound, that's quite different to the various types of kick used in other rock genres. Most obviously though is EDM (and most of it's sub genres), which evolved from the club scene, with their big sound systems and serious subs. DJ's started incorporating these sub into their compositions/performances and used them in much the same way as we do in film, to move masses of air to provide the audience with a real physical sensation. Large, live heavy metal gigs have been doing that for many years but with EDM it's use is much more creative, giving the audience a more normal range kick sound, teasing them and building up to the climax of the powerful sub supplemented kick sound. And because of the enormous popularity of EDM gigs, other artists and genres are now sometimes more inclined to incorporate some very LF synth material.

G
 
Last edited:
Nov 5, 2018 at 2:40 AM Post #10,311 of 14,606

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,548
Likes
4,462
Location
Hollywood USA
Mighty fancy!

My player automatically compensates for bass management with SACDs. I think most current players do. There's usually a setting to set speakers to "small" and/or to set the player to LFE or normal. If you set it to small and LFE, there is no need to adjust anything when you play SACDs. It's all transparent. If you have a sub, I don't see any reason to not engage bass management. It can only help. Maybe for quad... maybe, but i still prefer to have bass management on for quad.
 
Last edited:
Nov 5, 2018 at 9:54 AM Post #10,312 of 14,606

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,168
[1] P.S. 85+10 = 91? :confused:
[2] This is a much more relevant question to me because I mix down the sub with centre and recalculate sub output by crossing over my specified LF part of all channels to the sub. So the answer to the above is the number by which I should be amplifying the LFE channel before mixing it into the centre channel, correct? Does this apply to multichannel music as well?
[3] Would the requisite boost be already done by computer video player apps like say Netflix going out to a (virtual) sound card?

1. Yes, absolutely! This is the point I've tried to explain several times, apparently unsuccessfully. I'll try an be as absolutely clear as I can: The basis of all level calibration for 5.1, 7.1, etc., is the main/screen speakers. Everything else, the LFE/sub and surrounds are calibrated relative to that main/screen speaker calibrated level. The -3dB for the surrounds (cinema only) and the +10dB in-band gain for the LFE (cinema and home) are -3dB and +10dB relative to the screen/main centre speaker, which in the theatrical specification is 85dBSPL (C). And to be clear, this 85dBC is the level measured with an SPL meter with the main/screen speaker outputting 20Hz-20kHz pink noise (at -20dBFS). Our LFE is level calibrated to +10dB in-band gain relative to our main/screen speaker. The "In-band" part means that our LFE/Sub should be outputting 10dB more 20Hz-120Hz than the amount of 20-120Hz our main speaker is outputting. The issue should hopefully now be obvious, our main speaker is not outputting 85dBC of 20Hz-120Hz, it's outputting 85dBC of 20Hz-20kHz. If we remove the 121Hz-20kHz output of our main speaker (so that it's only outputting our required 20Hz-120Hz), it's output level will obviously be lower, it will be approx 81dBSPL and our sub is then calibrated 10dB higher, which is about 91dB. In practice, we wouldn't try and "remove the 121-20kHz", we'd just use an RTA to measure the 20Hz-120Hz portion of the main speaker's output. In other words, your equation should read: "85dBC - (the 121Hz -20kHz band) +10dB = 91dB".

2. This is where it can quickly get confusing, with your sub playing double duty, as I explained to Steve999 in the second paragraph of this post. To adapt the official manual LFE calibration procedure for a bass managed system, I would first sort out the bass management part of the equation. With say your centre speaker, use a sine sweep (or 20Hz-20kHz pink noise) and RTA software, and set the level of your sub so that you have as flat a response throughout the spectrum as practical, the main goal obviously is to equally balance both sides of the crossover point between your centre speaker and your sub. Once achieved, level calibrate your centre channel/sub combo to (say) 78dBC with 20Hz-20kHz pink noise at -20dB, using the "C" weighting, "slow" response on an SPL meter. Without changing that level (or the pink noise signal), use your RTA software (set to 1/3 octave bands) to take a 20Hz-120Hz plot of your centre/sub combo. Now take band-limited (compensated) 20Hz-120Hz pink noise, output that through your LFE channel (to your sub only) and take an RTA plot of that. Finally, EQ your LFE channel (NOT your sub!) so that each 1/3 octave band is 10dB higher than the corresponding band in the (20Hz-120Hz) plot you made of your centre/sub combo. However ...

3. I'm not sure but I would say probably not. If you try and achieve this by software in the digital domain, it has to be done the other way around. IE. You can't apply the boost to the LFE channel, you have to apply a reduction to all the other channels. Typically, you would accomplish the +10dB in-band LFE gain in the analogue domain. You can't just increase the LFE channel by the required amount in the digital domain. What would happen in the digital domain if the LFE channel on a particular film peaks at say -1dBFS and you try to add 10dB? As you can't have +9dBFS what you actually get is a nasty clip! Even if you were considering the calibration method I explained in point #2, that final 10dB per band LFE channel gain would have to be done with an analogue EQ or, you'd have to devise some way of applying it's inverse (reduction) to your main/screen speakers.

I know the above is confusing but again, it was never designed for consumers. In fact, it wasn't even designed for us sound engineers, it was designed by Dolby for the sole use of it's own technicians! Part of the licencing contract which allowed film mix stages to produce films in 5.1, 7.1, etc., required that Dolby's own technicians calibrated the mix stage and the mix/sound engineers weren't allowed to touch it! The problem with multi-channel music production is that it's relatively rare (outside of the film music world) and the music engineers sometimes have the same lack of understanding of LFE calibration, even some of the most respected engineers. So, even if your LFE is correctly calibrated, you maybe trying to reproduce music which isn't! As a general rule, for most people, I would advise that you let your AVR deal with it.

Thank you so much, Gregorio.

You're welcome!

My player automatically compensates for bass management with SACDs. I think most current players do.

If someone is outputting their SACD to an AVR though, wouldn't the AVR be boosting the LFE channel?

G
 
Last edited:
Nov 5, 2018 at 12:13 PM Post #10,313 of 14,606

Joe Bloggs

Member of the Trade: HiBy / EFO technologies Co
His Porta Corda walked the Green Mile
Joined
Oct 21, 2001
Posts
11,280
Likes
3,065
Location
Hong Kong and Melbourne
1. Yes, absolutely! This is the point I've tried to explain several times, apparently unsuccessfully. I'll try an be as absolutely clear as I can: The basis of all level calibration for 5.1, 7.1, etc., is the main/screen speakers. Everything else, the LFE/sub and surrounds are calibrated relative to that main/screen speaker calibrated level. The -3dB for the surrounds (cinema only) and the +10dB in-band gain for the LFE (cinema and home) are -3dB and +10dB relative to the screen/main centre speaker, which in the theatrical specification is 85dBSPL (C). And to be clear, this 85dBC is the level measured with an SPL meter with the main/screen speaker outputting 20Hz-20kHz pink noise (at -20dBFS). Our LFE is level calibrated to +10dB in-band gain relative to our main/screen speaker. The "In-band" part means that our LFE/Sub should be outputting 10dB more 20Hz-120Hz than the amount of 20-120Hz our main speaker is outputting. The issue should hopefully now be obvious, our main speaker is not outputting 85dBC of 20Hz-120Hz, it's outputting 85dBC of 20Hz-20kHz. If we remove the 121Hz-20kHz output of our main speaker (so that it's only outputting our required 20Hz-120Hz), it's output level will obviously be lower, it will be approx 81dBSPL and our sub is then calibrated 10dB higher, which is about 91dB. In practice, we wouldn't try and "remove the 121-20kHz", we'd just use an RTA to measure the 20Hz-120Hz portion of the main speaker's output. In other words, your equation should read: "85dBC - (the 121Hz -20kHz band) +10dB = 91dB".

2. This is where it can quickly get confusing, with your sub playing double duty, as I explained to Steve999 in the second paragraph of this post. To adapt the official manual LFE calibration procedure for a bass managed system, I would first sort out the bass management part of the equation. With say your centre speaker, use a sine sweep (or 20Hz-20kHz pink noise) and RTA software, and set the level of your sub so that you have as flat a response throughout the spectrum as practical, the main goal obviously is to equally balance both sides of the crossover point between your centre speaker and your sub. Once achieved, level calibrate your centre channel/sub combo to (say) 78dBC with 20Hz-20kHz pink noise at -20dB, using the "C" weighting, "slow" response on an SPL meter. Without changing that level (or the pink noise signal), use your RTA software (set to 1/3 octave bands) to take a 20Hz-120Hz plot of your centre/sub combo. Now take band-limited (compensated) 20Hz-120Hz pink noise, output that through your LFE channel (to your sub only) and take an RTA plot of that. Finally, EQ your LFE channel (NOT your sub!) so that each 1/3 octave band is 10dB higher than the corresponding band in the (20Hz-120Hz) plot you made of your centre/sub combo. However ...

3. I'm not sure but I would say probably not. If you try and achieve this by software in the digital domain, it has to be done the other way around. IE. You can't apply the boost to the LFE channel, you have to apply a reduction to all the other channels. Typically, you would accomplish the +10dB in-band LFE gain in the analogue domain. You can't just increase the LFE channel by the required amount in the digital domain. What would happen in the digital domain if the LFE channel on a particular film peaks at say -1dBFS and you try to add say 6dB? As you can't have +5dBFS what you actually get is a nasty clip! Even if you were considering the calibration method I explained in point #2, that final 10dB per band LFE channel gain would have to be done with an analogue EQ or, you'd have to devise some way of applying it's inverse (reduction) to your main/screen speakers.

I know the above is confusing but again, it was never designed for consumers. In fact, it wasn't even designed for us sound engineers, it was designed by Dolby for the sole use of it's own technicians! Part of the licencing contract which allowed film mix stages to produce films in 5.1, 7.1, etc., required that Dolby's own technicians calibrated the mix stage and the mix/sound engineers weren't allowed to touch it! The problem with multi-channel music production is that it's relatively rare (outside of the film music world) and the music engineers sometimes have the same lack of understanding of LFE calibration, even some of the most respected engineers. So, even if your LFE is correctly calibrated, you maybe trying to reproduce music which isn't! As a general rule, for most people, I would advise that you let your AVR deal with it.

Thanks greg. Re part 3, actually attenuation relative to max system volume is included in the VSTHost volume levels, so unless I were running the sound full blast digitally, I can do boosts just fine. The speakers on my system are set once to be able to play very loud at 0dBFS, without being so loud as to hiss noticeably, and volume control is all in software from there on out.
 
Nov 5, 2018 at 1:22 PM Post #10,314 of 14,606

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,168
Thanks greg. Re part 3, actually attenuation relative to max system volume is included in the VSTHost volume levels, so unless I were running the sound full blast digitally, I can do boosts just fine. The speakers on my system are set once to be able to play very loud at 0dBFS, without being so loud as to hiss noticeably, and volume control is all in software from there on out.

Yep, that would work. The only issue is that you're effectively loosing 10dB of your system's signal to noise ratio performance. That loss is probably a non-issue in many cases but worth baring in mind.

G
 
Last edited:
Nov 5, 2018 at 2:26 PM Post #10,315 of 14,606

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Posts
9,364
Likes
4,749
Yep, that would work. The only issue is that you're effectively loosing 10dB of your system's signal to noise ratio performance. That loss is probably a non-issue in many cases but worth baring in mind.

G
*castleofargh whistles as his reading of max true peak for the all session is -11.8dB(replaygain+EQ), and then foobar's volume removes another -7.35dB just for lazy loudness setting. SNR is so 2017 anyway. ^_^
 
Nov 6, 2018 at 12:46 PM Post #10,316 of 14,606

AKGForever

New Head-Fier
Joined
Oct 26, 2018
Posts
20
Likes
5
Location
New Jersey
Got kind of contradictory results on the digitalfeed.net ABX tests. I only did the minimum 5 tests as I was sure I wouldn't hear any difference. The first time I tried it I came up with 44% right, within the range of a coin flip. The only real outlier was the Eagles' Hotel California, which was the only song I was actually familiar with. I got 80% right on Hotel California. Didn't think too much about as if you flip a coin 25 times, it might come up heads four times in a row. I did the test again a day later and scored 52% but got 100% on Hotel California. Mostly I was guessing and I can't point to anything that was a give away on Hotel California. Just found it curious. BTW, this was using a Windows laptop with "HD" sound and AKG K240 headphones.
 
Last edited:
Nov 6, 2018 at 1:34 PM Post #10,317 of 14,606

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,548
Likes
4,462
Location
Hollywood USA
f someone is outputting their SACD to an AVR though, wouldn't the AVR be boosting the LFE channel?

It all evens out. There's no reason for regular folk to worry about it. You put in a disc and it plays the way it's supposed to.
 
Nov 6, 2018 at 3:37 PM Post #10,318 of 14,606

KeithEmo

Member of the Trade: Emotiva
Joined
Aug 13, 2014
Posts
1,696
Likes
856
Interesting......

I've always found the HDTracks 24/192k version of that album to be excellent for listening for minor differences between DACs and filter choices.
(Listen for slight differences in the sound of the plucked guitars and cymbals.)

Got kind of contradictory results on the digitalfeed.net ABX tests. I only did the minimum 5 tests as I was sure I wouldn't hear any difference. The first time I tried it I came up with 44% right, within the range of a coin flip. The only real outlier was the Eagles' Hotel California, which was the only song I was actually familiar with. I got 80% right on Hotel California. Didn't think too much about as if you flip a coin 25 times, it might come up heads four times in a row. I did the test again a day later and scored 52% but got 100% on Hotel California. Mostly I was guessing and I can't point to anything that was a give away on Hotel California. Just found it curious. BTW, this was using a Windows laptop with "HD" sound and AKG K240 headphones.
 
Nov 6, 2018 at 3:52 PM Post #10,319 of 14,606

TheSonicTruth

1000+ Head-Fier
Joined
Dec 19, 2014
Posts
1,018
Likes
126
Interesting......

I've always found the HDTracks 24/192k version of that album to be excellent for listening for minor differences between DACs and filter choices.
(Listen for slight differences in the sound of the plucked guitars and cymbals.)

More likely to hear mastering differences than differences between DACs even at that resolution.
 
Nov 6, 2018 at 3:55 PM Post #10,320 of 14,606

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,548
Likes
4,462
Location
Hollywood USA
HD Tracks downloads often have no super audible content except for spikes of complete noise. I posted numerous examples of that a few months back. You might want to skip pages 2 through 7 in that thread though. It gets really good again at pages 8 and 9.

https://www.head-fi.org/threads/what-kind-of-ultrasonic-frequencies-are-in-hd-tracks.885484/

There's absolutely no guarantee that HD Tracks downloads are any better than the commercial CD, and in many cases, it's clearly worse.
 
Last edited:

Users who are viewing this thread

Top