Why do so many recording sound bright or "harsh"?

[Dimifoot]

What you describe as harshness could be an issue between 2000-8000Hz.
Lots of reflective surfaces also could be faulted.

 

[Robin L]

Yes I have at least in a crude way. I have used my Dayton OmniMic system to measure my speakers near-field and away from reflecting surfaces; I have also measure the response from my listening position.

My room does have a major effect but it isn't to exaggerate the highs, if anything it rolls them off a bit. However there are mid-range bands where the response is cut or boosted. However my current strategy is to leave the mid-range unadjusted -- my speakers are pretty flat as measured in the near field. I think approach gives me the most transparency and 'air', (though maybe I'm self-deluded).

If you're only really noticing the problem on a few recordings, then it's because the recordings you dislike are having their problems exposed.

 

[pozz]

Very interesting. For my part I listen to mainly Classical music where compression and high mix levels are used to a much less extent. Nevertheless "harshness" is very common; it's a notable problem for violin, brass, soprano voices, and the like. The best recordings I have, at least for large scale works, the mix levels are very low compared to Jazz or popular music recordings, or even chamber music recordings.

I think that's a different story then. In classical music I would guess that issue is mic placement. Radiation patterns for instruments are really variable. It's not my kind of music so I don't have too many examples but one I know of is using overhead mics for violins is liable to make them sound crispy.

I can't find the research I was thinking of. Below are just horizontal planars. For 2kHz, you can bet the energy that is directed horizontally at lower frequencies is oriented more vertically.

 

[Feanor]

How old are you? Age-related hearing issues could play a part. I seem to be a little more sensitive to some frequencies than I was in the past.

I'm very aged. Equally to the point I'm basically stone deaf above 10kHz.

 

[Feanor]

If you're only really noticing the problem on a few recordings, then it's because the recordings you dislike are having their problems exposed.

Well I notice the problem on a good many of my recordings. But for sure, my system exposes recordings with the problem.

 

[ahofer]

It's an interesting exercise to search Beethoven 7th Symphony in Tidal, and sample through each recording. I recommend about 4-5 minutes into the first movement, a nice tutti presentation of the most memorable theme, for comparison. Lots of approaches, many horrid recordings. The old ones hold up pretty well.

 

[wemist01]

Honestly I think its your room/gear. I don't have these issues. And I am very nitpicky about tone and hate treble issues.

I don't think the claim carries weight without examples. I'd also like to square the original poster's age, and his level of hearing loss with the assertion. If the poster is as old as he looks and has negligible hearing loss, I'm astounded. I'm 51 and will need hearing aids soon- no joke. So this isn't something I notice.. most of the time. The Moinet/Erhardt Berlin Classics disc of Bach concertos from last year is painful (someone please comfirm!)

 

[sigbergaudio]

Would be interesting with a couple of examples of tracks that are bright, so we could have a listen (preferrably something available on Spotify or something so everyone could easily access it). Not sure if it would help us determine the cause 100%, some may for various reasons think it sounds bright or not, but I for one would be curious to test on my own system.

Perhaps you @Feanor could also elaborate a bit on your definition of bright. Is it immediately harsh to your ears when you start the track, or is it more about listening fatigue over time, etc?

 

[Sgt. Ear Ache]

I'm very aged. Equally to the point I'm basically stone deaf above 10kHz.

yes but harshness could be occurring in the 2-5khz range...

 

[Robin L]

I don't think the claim carries weight without examples. I'd also like to square the original poster's age, and his level of hearing loss with the assertion. If the poster is as old as he looks and has negligible hearing loss, I'm astounded. I'm 51 and will need hearing aids soon- no joke. So this isn't something I notice.. most of the time. The Moinet/Erhardt Berlin Classics disc of Bach concertos from last year is painful (someone please comfirm!)

Listening to the YouTube version of this, a little bright, but nothing crazy.
You would probably like this more:

 

[ahofer]

Here's a good contrast. Start the "Mozart 225" collection of symphonies from DG/Decca (first symphony is English Concert/Trevor Pinnock),then compare to the first parts of the Philips complete Mozart symphonies (Neville Mariner/Academy of St. Martin in the Fields 1995).

or search for "Mozart Symphony No. 1 in E flat Allegro" and compare those and other tracks.

The DG/Deccca sounds pretty harsh, at least thru my Sennheiser's now. Of course, the first thing you have to get used to is the pitch difference (oy- Berlin Philharmonic). YOW!

 

[ElNino]

Very interesting. For my part I listen to mainly Classical music where compression and high mix levels are used to a much less extent. Nevertheless "harshness" is very common; it's a notable problem for violin, brass, soprano voices, and the like.

My room does have a major effect but it isn't to exaggerate the highs, if anything it rolls them off a bit. However there are mid-range bands where the response is cut or boosted.

Based on these two posts, I think it might be related to the way your speakers interact with your room in the midrange. If brass instruments are harsh, that's not really harsh "highs" per se, though the brain perceives it as such... most trumpet content is actually in the 300-1kHz range (typically the overtones aren't what cause harshness).

One thing I've found helpful in the past is running a real-time analyzer while listening to music. This basically lets you see the frequency bands that are loud when you hear harshness. At least for me, I realized a lot of what I thought of as harsh highs were actually anomalies under 2kHz.

 

[richard12511]

I increasingly find the opposite to be true, almost no modern recordings sound bright on my system. I find that many older recordings (80s and earlier) are lacking in bass, and may sound "bright" as a result, but not necessarily because they have too much treble. Depends a bit on what genres you listen to as well I guess. Some rock genres (and periods) tend to be pretty bright.

I also don't agree that compression / low dynamic range necessarily must mean harsh sound.

Note that I have a system that is intentionally emphasizing the low end, not a flat response.

Speculation: If you find many recordings to be too bright, the balance may be a bit off in your system - do you simply have loudspeakers that is too bright for your liking? Or possibly the room have too many hard surfaces, resulting in excessive reflections and too much information in the top end as a result.

Agreed. It's a shame, too, as a lot of the music that I really enjoy is older stuff that just sounds so harsh(relative to modern music).

 

[Robin L]

Here's a good contrast. Start the "Mozart 225" collection of symphonies from DG/Decca (first symphony is English Concert/Trevor Pinnock),then compare to the first parts of the Philips complete Mozart symphonies (Neville Mariner/Academy of St. Martin in the Fields 1995).

or search for "Mozart Symphony No. 1 in E flat Allegro" and compare those and other tracks.

The DG/Deccca sounds pretty harsh, at least thru my Sennheiser's now. Of course, the first thing you have to get used to is the pitch difference (oy- Berlin Philharmonic). YOW!

But that's not simply "harsh recordings". Historically informed performances on original instruments tend to be brighter/harsher, don't use vibrato [or not as much], have other reasons why they sound more harsh. Also, HIP performances usually use A=415 instead of A=440.

 

[LuvTheMusic]

One other issue that can contribute to bright or aggressive sound for upper strings (violins) in orchestral recordings: frequently, recording mics are suspended high in front of or even over the string sections. This means that the sound they are recording is the sound as projected from directly the top of the instruments. In contrast, when you are sitting in a concert hall, the sound that you are hearing is from the side (presumably a mix of sound radiating from the top and reflected to you and sound radiated from the sides and even the bottom of the instruments.

If you have a chance to hear the sound of a violin from above or with the instrument tilted almost 90 degrees toward you, you will find that the sound is considerably brighter than what you hear normally. (You may not have noticed this previously because most of us don't have much opportunity to listen to violins from above, e.g., while perched on ladders!)

Of course, the recording engineer could move the mics, could adjust for the brightened sound, etc. But none of that may happen.

 

[RayDunzl]

One thing I've found helpful in the past is running a real-time analyzer while listening to music.

Looking at the recording can be informative, too.

Left channel WAV file (from above) dropped onto the RTA window in REW:

This curve is typical.

I find "brightness" dependent upon where the slope in the higher frequencies begins.

The above looks good, roll starting at 1.5kHz...

---

Here's a recording (from CD) that I plan to "remaster" with a different EQ - it sounds rather "thin".

Carla Bley - European Tour - Rose and Sad Song

Here, the brightness is more due to a lack of energy in the lower frequencies, despite the highs being reasonably rolled.

Lastly a highly engineered (won a Grammy for it) modern pop recording:

Daft Punk - Random Access Memories - Give Life Back to Music

No lack of low-end there.

 

[raindance]

I have used and still use so roll-off of highest frequencies -- it helps a little but rarely gets rid of problem. That is, I think there is something more to it than merely an upward sloping response curve on the recording.

Brightness is not highest frequencies. It's in the 2-6KHz range. Tone controls usually have a 10KHz corner frequency. And I agree with the original poster that a lot of recordings are too bright. I think it's because of mastering engineers who don't give a shit. I have a Quad preamp with useful tone controls, including spectral tilt.

 

[Robin L]

Brightness is not highest frequencies. It's in the 2-6KHz range. Tone controls usually have a 10KHz corner frequency. And I agree with the original poster that a lot of recordings are too bright. I think it's because of mastering engineers who don't give a shit. I have a Quad preamp with useful tone controls, including spectral tilt.

Two things, constant exposure to loud sound makes sound engineers ears burn out faster, and some engineers I've worked with used mellow sounding monitors [so they don't burn out their ears], resulting in recordings that are more harsh sounding on gear that's more open on top.

 

[Feanor]

yes but harshness could be occurring in the 2-5khz range...

Since I can hear it, apparently it is.

 

[Feanor]

Brightness is not highest frequencies. It's in the 2-6KHz range. Tone controls usually have a 10KHz corner frequency. And I agree with the original poster that a lot of recordings are too bright. I think it's because of mastering engineers who don't give a shit. I have a Quad preamp with useful tone controls, including spectral tilt.

Thank you for pointing out the the harshness problem tends to be in the 2-6kHz range. This is something that I might considering when hoping to use EQ to reduce brightness/harshness.