1/5 of a dB difference

[pucho812]

Was out at a place today where the claim is left side sounds louder and different frequency response vs the right side of one of several speaker monitor paths.

In a room where noticeable comb filtering is happening just by leaning forward or backwards at the mix position I have doubts.

So to eliminate the room+speakers, out came the AP and did testing. I measure an average of 4.1dBu on the right and 4.3 dBu on the left in terms of level. Without level changes the frequency responses could trace atop each other. It’s the same curve. I can’t imagine most folks can hear that.
That’s 1/5th a dBu
Be honest, how many you think can hear that? Or better yet, let’s poll and see yes you can or no you can’t.
I know I can’t.

 

[iomegaman]

Nope

 

[Gold]

It is highly dependent on the source audio. On a solo instrument, no way. On a dense mix, yes way. No golden ears necessary.

 

[JohnRoberts]

Was out at a place today where the claim is left side sounds louder and different frequency response vs the right side of one of several speaker monitor paths.

In a room where noticeable comb filtering is happening just by leaning forward or backwards at the mix position I have doubts.

So to eliminate the room+speakers, out came the AP and did testing. I measure an average of 4.1dBu on the right and 4.3 dBu on the left in terms of level. Without level changes the frequency responses could trace atop each other. It’s the same curve. I can’t imagine most folks can hear that.
That’s 1/5th a dBu

1/5th dB difference. dBu is an absolute level.

Be honest, how many you think can hear that? Or better yet, let’s poll and see yes you can or no you can’t.
I know I can’t.

I don't argue with people about what they say they can hear, especially on the internets. Proof involves rigorous double blind testing with results that reach statistical significance. Are you getting paid by the hour to do that?

JR

 

[pucho812]

It is highly dependent on the source audio. On a solo instrument, no way. On a dense mix, yes way. No golden ears necessary.

You sir, As a trained mastering engineer with loads of experience I don’t doubt it. But the average tracking engineer is highly suspect.

 

[pucho812]

1/5th dB difference. dBu is an absolute level.

I don't argue with people about what they say they can hear, especially on the internets. Proof involves rigorous double blind testing with results that reach statistical significance. Are you getting paid by the hour to do that?

JR

Sorry JR got my wording cornfused. Yes the difference was .2dBu so about 1/5 a dBu

 

[Gold]

You sir, As a trained mastering engineer with loads of experience I don’t doubt it. But the average tracking engineer is highly suspect.

The average tracking engineer is dealing with individual instruments not completed mixes. I agree it’s unlikely to make a big difference in a tracking room. The more dense the mix the easier it is to hear small level changes. There is no hard threshold.

 

[JohnRoberts]

Sorry JR got my wording cornfused. Yes the difference was .2dBu so about 1/5 a dBu

I will apologize in advance for being pedantic (precise) but dB (decibels) are units of relative power (often used for voltage ratios assuming constant impedance termination). dBu is a voltage level relative to 0.775V AC.

[TMI: dBV is voltage level relative to 1 Vac, dBm is power level relative to 1 mW /TMI]

I believe you are trying to express a .2 or 1/5 of decibel difference, a small difference generally considered inaudible.

JR

 

[sodderboy]

I cannot, but I never question any working engineer regarding what they hear.
Some of my memorable teachable tech moments involve hound doggin’ L-R differences all around studios, and the solutions almost always involved capacitance irregularities in components, wire/cable, PCBs, anything passing current.
And I DMM cal to the hundredth of a dB, because I can. Zero is zero.
Mike

 

[user 37518]

The solution is simple: Swap the monitors and measure again. If the difference remains the same, it's the room and not the monitors. If the difference swaps sides, trim the offending monitor -0.2 dB.

P.S. A rule in acoustic measurements is that one single measurement means nothing, make several measurements changing the mic position just a bit (I am talking about a few cm) or maybe rotate it just a tad, take the average of all of the measurements (sum all the results and divide by the total number of measurements), and then check if there is a difference.

P.P.S. I've heard stories of techs servicing studios where the engineers boast about the lovely sound of their bazillion dollars, ultra atomic quantum radioactive Word Clock, only to notice that the computer has the clock source set to 'Internal'

 

[Ó Draighnáin]

The average tracking engineer is dealing with individual instruments not completed mixes. I agree it’s unlikely to make a big difference in a tracking room. The more dense the mix the easier it is to hear small level changes. There is no hard threshold.

I wouldn't be able to hear the difference. I also do not believe that I even reach the skill threshold required for "average tracking engineer", I'm not in the industry. However, I do recall that at times during mixing of an album, the movement of a single stem level only a fraction of a decibel appeared make a audible difference in respect of that particular stem within the mix?

 

[pucho812]

The solution is simple: Swap the monitors and measure again. If the difference remains the same, it's the room and not the monitors. If the difference swaps sides, trim the offending monitor -0.2 dB.

P.S. A rule in acoustic measurements is that one single measurement means nothing, make several measurements changing the mic position just a bit (I am talking about a few cm) or maybe rotate it just a tad, take the average of all of the measurements (sum all the results and divide by the total number of measurements), and then check if there is a difference.

P.P.S. I've heard stories of techs servicing studios where the engineers boast about the lovely sound of their bazillion dollars, ultra atomic quantum radioactive Word Clock, only to notice that the computer has the clock source set to 'Internal'

Yes..I swapped and proved what the desk is doing. The fact the guy doesn’t hear it as the ap says it is, is different. I am not concerned with acoustics here as I do not need to solve those issues. But I can point it out. I did notice how much of a difference leaning forward and backwards changed the audio while at the mix position. Nothing I would want in a room. The main focus as it relates to me is the gear. I trust the ap as far as what is coming out of the desk before it hits amps and speakers. Anything after the desk is beyond the job scope but I can give advice.

 

[Gold]

Everyone who says they can’t hear that small a change should try listening to a dense mix and changing the level by 0.1dB. Something like NIN is a good example. You may be surprised.

 

[user 37518]

Everyone who says they can’t hear that small a change should try listening to a dense mix and changing the level by 0.1dB. Something like NIN is a good example. You may be surprised.

In addition to that, it would be good for someone else to do it for you while you are not looking. As to not create a "Producer's Fader" effect.

 

[Gold]

In addition to that, it would be good for someone else to do it for you while you are not looking. As to not create a "Producer's Fader" effect.

Always better to do it that way. I’ve sat with many engineers over the years doing album sequencing. So I’ve been the one to change the setting. They can hear it, I can hear it. It’s repeatable.

I find as long as you are aware of expectation bias it’s avoidable. If something is close or I’m mot sure I make note of that. It may or may not be. If I clearly hear a difference or clearly don’t hear a difference I’m confident in that assessment.

 

[JohnRoberts]

I will repeat that I do not dispute what other people say they can hear on the WWW. My personal experience wrt listening level comparisons was mostly last century between similar power amps with slightly different output power levels. This is not exactly apples to oranges with near field listening over studio monitors.

I often operated these listening tests single blind, with unskilled listeners. The simple power amps comparisons were sometimes conflicted by things like (clip limiting). For an unfortunate factoid, amplifiers allowed to clip are perceived as louder than the same power amps operated cleanly. In fact amplifiers allowed to clip can put out more average power, than a clean clip limited path. I have even experienced smaller amps that were considered more powerful (louder) than the larger amps when allowed to clip.

Sorry if this is too much of a veer.

JR

 

[user 37518]

I will repeat that I do not dispute what other people say they can hear on the WWW. My personal experience wrt listening level comparisons was mostly last century between similar power amps with slightly different output power levels. This is not exactly apples to oranges with near field listening over studio monitors.

I often operated these listening tests single blind, with unskilled listeners. The simple power amps comparisons were sometimes conflicted by things like (clip limiting). For an unfortunate factoid, amplifiers allowed to clip are perceived as louder than the same power amps operated cleanly. In fact amplifiers allowed to clip can put out more average power, than a clean clip limited path. I have even experienced smaller amps that were considered more powerful (louder) than the larger amps when allowed to clip.

Sorry if this is too much of a veer.

JR

With that discussion about amplifiers you are opening a can of worms, John, or rather Pandora's box...

 

[Gold]

I often operated these listening tests single blind, with unskilled listeners. The simple power amps comparisons were sometimes conflicted by things like (clip limiting).

Power Amps are hard because the only way to test them is subjectively. An electronic load can only tell you so much. Many power amps will sound very similar until you have them drive a complicated load. Then they can sound not similar at all.

 

[cyrano]

As far as I can tell, a room that has comb filtering goin' on, isn't suitable to hear such a minute difference, even on a mono mix.

I would be extremely amazed if the comb filtering was symmetrical.

 

[JohnRoberts]

Power Amps are hard because the only way to test them is subjectively. An electronic load can only tell you so much. Many power amps will sound very similar until you have them drive a complicated load. Then they can sound not similar at all.

There are any number of valid objective performance metrics for audio power amplifiers. As product manager for all of Peavey's powers amps for a time, and having designed one myself indeed there are unobvious deviations from ideal that casual listeners don't recognize. As the popular wisdom goes, audio amplifiers all sound alike when operated in their linear*** region.

The most common source of misdiagnosed audible faults is current limiting. There are sundry different protection schemes ranging from simple current limiting, to VI limiting, and even secondary breakdown protection (for bipolar power transistors). This can get pretty hairy on the audiophile side because I have seen some consumer speaker crossovers that involved crazy frequency dependent loads. The worst examples I saw were from esoteric speaker companies who didn't also manufacture amplifiers. At Peavey we got pretty immediate feedback if a new Peavey speaker didn't play nice with the Peavey amplifiers. Then we have customers running amplifier loads down to 2 ohms to squeeze out a little extra power.

Audible amplifier current limiting is hard to diagnose because it is not a steady state phenomenon but varies with the music. Slew limiting is probably over stated as a potential audible fault, but again I have seen poorly designed amplifiers with asymmetrical slew limiting that would introduce DC into the unlucky speaker. I recall at Peavey when we had to replace too many speakers that were killed by a competitors poorly designed power amp. They had a better reputation than Peavey so customers assumed the malfunctioning amp was just powerful for the Peavey speakers. Eventually the competitor figured out what was wrong with their amp and fixed the design.

Another potentially subtle audible fault is instability during clip events. We've seen bursts of HF oscillation when some poorly designed amplifiers clip. Some even see DC errors related to that instability.

These are just a few of a much longer list. I like to joke that power amps are the most complicated simple design, right after mixers.

JR




***Linear region, means not clipping, not slew limiting, not current limiting, not doing anything other than a linear version of the input signal with gain.