John Hammond / Rough & Tough / Insane Treble

[MRC01]

John Hammond, Rough & Tough, Chesky Records, 2009.

Listening to this recording on Qobuz, the treble at first sounded a bit bright, then the harmonica hit and my ears just about exploded. There's some kind of powerful resonance at very high frequencies, sounds like a square wave, almost like a 19+20 kHz IMD test tone.

For the full effect, listen to track 5 when the harmonica enters around 0:58. But first brace yourself! I tried applying the CD de-emphasis curve (even though it's a 192-24 recording), but that flattened the lower treble and still wasn't enough at high frequencies.

But, nobody mentions this in album reviews. They think it sounds great. This makes me wonder: are they deaf, or is this a flaw in the Qobuz release of this recording?

 

[Elkerton]

I can’t attest to your claim of the extremely high resonance. I’m old, but I’m listening to it as I type in multichannel from a SACD, and it sounds just fine. It’s has only 4 channels because, I’m guessing, it was recorded with a Calrec mic which has only 4 capsules in order to catch a single point in space. It’s a keeper.

 

[kemmler3D]

I might be deaf because I'm not hearing anything TOO weird over here on Tidal, other than some very pronounced harmonica.

There's some kind of powerful resonance at very high frequencies, sounds like a square wave, almost like a 19+20 kHz IMD test tone.

Could it be breakup or distortion on your system?

Either way, I am hearing pretty pronounced pumping effects from whatever compressor they used, *from the harmonica*, which is insane, you almost never hear lows or mids ducking highs. So you have a very valid point that the treble is mixed too hot.

e: as the next post shows you have 4khz coming in at the same level as 100hz, that's not normal.

e2: I guess they recorded everything on the one mic and there was just not much to be done about the harmonica being objectively way louder than everything else at that distance? I still would have tried harder, with dynamic EQ and multiband compression and stuff.

e3: It actually is sort of like a square wave, in that it has a long series of harmonics at almost equal levels... lol, good ear @MRC01

 

[MRC01]

Your reply reminded me to analyze the wave. Here's the part I mentioned in track 5 where the harmonica enters and the high frequency blast starts:

The green shows the frequency content just before the harmonica enters. The purple part is taken a few seconds later while the harmonica plays. No question, there's a huge amount of extreme HF energy there. No harmonica I've heard in real life sounds like that, it's definitely something with the recording.

PS: distortion on my system measures -60 dB and lower from 2 kHz on up (Magnepan 3.6/R tweeters are very clean).

 

[MRC01]

... e2: I guess they recorded everything on the one mic and there was just not much to be done about the harmonica being objectively way louder than everything else at that distance? I still would have tried harder, with dynamic EQ and multiband compression and stuff.

The liner notes say they used a special binaural mic, some kind of single multi-capsule unit.

 

[Elkerton]

The Calrec mic (Calrec may not be who makes the mic now, but it is originally their design and manufacture. The notes call it a Soundfield mic, which Calrec also referred to the design as such) has 4 diaphragms in one capsule, one facing up front, another facing down rear; one facing down front, the other facing up rear. Chesky uses no compression and a very simple equipment chain.

 

[MRC01]

Detailed info on this mic seems hard to find, but I found this document that shows how its frequency response varies with orientation and how to calibrate it. Some of the arrangements (such as on slide 84, WXY at 45 degrees) show big frequency response peaks at 7k and 16k.

Maybe that's what happened with this recording? Whatever the cause, it surprises me that Chesky released it sounding like this. There clearly is something wrong with the way the harmonica sounds - distorted and painful to listen to.

The other possibility is that the version of this recording on Qobuz is flawed and doesn't match the original SACD release.

 

[kemmler3D]

Chesky uses no compression and a very simple equipment chain.

Simple equipment chain I would believe, but there's definitely a lot of compression / limiting on this recording. You can hear it plain as day as soon as the harmonica comes in. In this case I don't think they had any choice. Seems like the harmonica was just like, 12dB louder than anything else the mic picked up. If they shipped the mix in its raw state, it would be completely unlistenable.

I think it's probable that they don't usually use the same mic for vocals and harmonica, but I've never recorded harmonica.

 

[Kal Rubinson]

I listened to the original 4 channel SACD rip and, yes, that harmonica is obtrusively loud at that point and at 1 or 2 other points in the track. In between, it's OK, so it seems that they have done some spotlighting at those points. OTOH, even at those points, I hear it pretty as clean but just too loud.

I have not tried it on Qouz.

 

[MRC01]

I listened to the original 4 channel SACD rip and, yes, that harmonica is obtrusively loud at that point and at 1 or 2 other points in the track. In between, it's OK, so it seems that they have done some spotlighting at those points. OTOH, even at those points, I hear it pretty as clean but just too loud.

I have not tried it on Qouz.

It would be interesting to see a spectrum analysis from the SACD to compare with the track from Qobuz. If anyone wants to do that, here are details on the snippets I posted above, both from track 5 of the 192-24 PCM on Qobuz:

Audacity 3.7.5, Analyze/Plot Spectrum, Hann Window, size 65536, log frequency. Green: 0:51 to 0:54; Purple: 0:57 - 1:00.

 

[Elkerton]

Looking at the frequency response of the mic in the link, it would have been beneficial if some PEQ had been applied on the peaks.

 

[Mnyb]

I have the HD tracks download , just very loud harmonica and i think closely miked , youd pick up stuff a listener would not normally hear at distance ?
it would likely sound like that if one played it into your ear ?

Microphones are not ears everything need some kind of post production .

 

[MRC01]

Looking at the frequency response of the mic in the link, it would have been beneficial if some PEQ had been applied on the peaks.

If I understand the graphs, they measured FR at different mic orientations to the source. This changes the response significantly: some are linear/flat, others are peaky/bumpy. If the mic was close to the source (as suggested in photos in the album liner notes), then the musician (Hammond) moving around while playing might change the orientation enough to shift the frequency response.

Maybe that is what happened at the spots where the extreme HF grabbed my attention.

 

[MRC01]

According to this, harmonica naturally has extreme HF harmonics. If so, it's similar in this regard to other instruments and sounds like bagpipes or jangling keys. If someone were playing a harmonica live right in front of you, these super high frequencies would not dominate the sound because the highest of these harmonics would be naturally attenuated by even just a few feet of air/space between you and musician. But if it is already a close-miced recording and then the musician leans forward toward the mic while playing loudly, the mic will capture these extreme frequencies and may even overload.

I've noticed this playing music in local groups. Sitting on stage right next to the violin (or any other instrument), it sounds way more edgy and bright than it does just a few feet away, and doesn't sound like a "natural violin" until you move further away to the 1st row of audience seats. No microphones or amps, this is about acoustic sound in natural spaces.

That may explain why the harmonica in this recording (at least in the moments mentioned above) sounds like an artificial caricature. Though even if it happened naturally, it seems to me like a preventable error in their mic positioning/setup. It's disappointing because I do like this album, both the music and the recording, except for the moments where it suddenly explodes in my ears.

This recording may need a warning similar to Stereophile Test CD #2 (all caps in original), "THESE TRACKS CONTAIN HIGH LEVELS OF HIGH FREQUENCIES AND MUST NOT BE PLAYED AT HIGH VOLUMES IF LOUDSPEAKER DAMAGE IS TO BE AVOIDED".