Dynamic Range...

[Covenanter]

Radio is still compressed of course. Classic FM do it on all their broadcasts I believe and even Radio 3 is sometimes compressed.

Chris

 

[The_Lhc]

Some records sound better with a low dynamic range. Phil Spector, Motown singles, stuff like that. :O)

Well, when you name your production technique the "Wall of Sound" you can't exactly expect subtlety...

 

[andyjm]

Radio is still compressed of course. Classic FM do it on all their broadcasts I believe and even Radio 3 is sometimes compressed.

Chris

FM radio has limited bandwidth and dynamic range. Classical music in particular has to be compressed as the transmission chain can't cope with it.

If you google 'what happens to the proms after the royal albert hall' there is BBC engineering article that explains the various steps.

Why can't I just cut and paste a link into this forum? I can on every other forum I frequent. Have Haymarket bought the 'home and student' version?

 

[Glacialpath]

Dynamic range has nothing to do the quality of a recording. For instance a Band like Napalm Death that play a pretty much full on noise (Grindcore) so there will be next to no dynamic range in their music as all the notes are played a basically the same loud volume.

Now capturing this and making a good recording of it, when played back will just sound loud. but the quality of the sounds made by the band will sound really good.

Now limit the sound wave to allow a louder volume when puting the audio on disc and you squash what little natural dynamics there are. An example being when a song breaks down to 1 guitar before the rest of the band kick back in, that one guitar in the real world would be quiter than when the whole band play. With the limiting however when the rest of the band kicks in the volume will stay the same. This is bad, undesirable and is what the loudness war is destroying.

Another example. I'm a drummer. The loudest instrument in the band without being miked up. If the band is playing a slow melodic passage with clean guitar sounds and the vocalist is not screaming. I won't be hitting as hard as I will when we kick in to a full on heavy bit. This would be dynamic range but the mastering process would eliminat the dynamic range we played if not done correctly.

So a Napalm Death Album before limiting will have a small dynamic range anyway so would score low on the DR web site but wouldn't indicate it was a bad recording as I said because the music is just loud anyway. It's a shame that it's not the case and the last 7 albums have fallen foul to the over limited mastering which makes them sound all flappy and harsh.

 

[lindsayt]

Napalm Death's 1990 release of Harmony Corruption has a DR of 14. Which is pretty good.

Compare this against their 2009 Time Waits for No Slave album with its DR of 5 to see which you think sounds better / more dynamic.

 

[Thompsonuxb]

Which sounds better?

 

[pauln]

I think that if one were to take two examples of the same recording, let's say an original CD from the late 80's with a high DR value and a remastered, compressed version from recent years with a low DR value, the former will sound better on a good quality system (headphones in my case) and the latter will sound better in a car or on lesser quality speakers for reasons already given. I think it's pretty straightforward really. I've got some music that I can't listen to on headphones, it hurts my ears, but in the car it sounds fantastic.

Both flavours should be available - it can't be that hard.

I don't think that's true.

Quality CD's sound like quality CD's regardless.

I do alot of MixCD's for my personal use if they sound good on my set it sounds good in the car.

compressed stuff sounds compressed and not enjoyable. but then again I'm comparing with what I know well recorded stuff 'can' sound like.

Is English your native language? It's just that sometimes you don't seem to understand what I write. Maybe it's me.

 

[BigH]

Dynamic range has nothing to do the quality of a recording. For instance a Band like Napalm Death that play a pretty much full on noise (Grindcore) so there will be next to no dynamic range in their music as all the notes are played a basically the same loud volume.

Now capturing this and making a good recording of it, when played back will just sound loud. but the quality of the sounds made by the band will sound really good.

Now limit the sound wave to allow a louder volume when puting the audio on disc and you squash what little natural dynamics there are. An example being when a song breaks down to 1 guitar before the rest of the band kick back in, that one guitar in the real world would be quiter than when the whole band play. With the limiting however when the rest of the band kicks in the volume will stay the same. This is bad, undesirable and is what the loudness war is destroying.

Another example. I'm a drummer. The loudest instrument in the band without being miked up. If the band is playing a slow melodic passage with clean guitar sounds and the vocalist is not screaming. I won't be hitting as hard as I will when we kick in to a full on heavy bit. This would be dynamic range but the mastering process would eliminat the dynamic range we played if not done correctly.

So a Napalm Death Album before limiting will have a small dynamic range anyway so would score low on the DR web site but wouldn't indicate it was a bad recording as I said because the music is just loud anyway. It's a shame that it's not the case and the last 7 albums have fallen foul to the over limited mastering which makes them sound all flappy and harsh.

Actually Napalm Death scores fairly well on DR with scores around 8-14, which was a surprise to me. But generally I agree with your comment, I feel some music is OK compressed while other genres fair less well.

 

[BigH]

On how many recordings that score 15 or more on the DR database have they mucked up other aspects of the recording, such as the clarity, focus, pitch, the mix - to the extent that it spoils the enjoyment of the music?

I haven't come across any.

If they've taken the care to preserve the dynamic range, it stands to reason they'd take good care of the other aspects of the recording.

For sure the natural dynamic range of music will depend upon the instrumentation and the type of track that it is.

But the DR in the DR database is a pretty good guide to how dynamic the recording will sound. And dynamics are vitally important in making a recording sound like actual music and less like a recording replayed through a sound system.

There's a game that we can all play. Listen to an album. Guess the DR. Look it up in the DR database to see how close you were.

But how many cds have a DR rating of 15 or more, must be less than 1%? Anyway if you take Kind of Blue the first cd was not so good but its 15, the 2009 remaster is much better and correct pitch as the wrong speed has been corrected but its only DR 13. The Gold version I can't listen to its too bright, hurts my ears, don't know the DR rating as its not on there, but probably is 14 or 15.

One surprise is Eva Cassidy Live at Blues Alley is only 9 thought that would be higher but I suppose there are not any really loud sounds?

Any reason why a lot of SACDs have lower DR than the redbook cds? I would expect them to have higher DR.

 

[davedotco]

Again I feel I have to chip in here.

Many of the complaints that I see in these debates (dynamic range, compression) is not about compression as such, not even over compression but about peak limiting.

We have all seen the traces of recordings where all the music peaks are pushed up hard to the 0dB line and this is, I think, the major problem. With the peaks 'cut off' in this manner we are virtually dealing with a series of square waves, which invariably bring a certain harshness to the recordings, a harshness that is particularly unpleasant on better equipment.

Sure, some compression techniques are pretty unpleasant too (side chain compression for example) but on most recordings it is the squaring off of overdriven peaks that causes the bigest problem.

 

[lindsayt]

BigH for Kind of Blue my preference would be the original stereo vinyl version or the 2 track tape. For the tracks recorded with a sharp pitch I'd just adjust the speed down a bit when I played them.

I'd want to listen to the various CD versions, level matched, on my own system before deciding which version I liked the best.

There's a high chance that Eva Cassidy Live at Blues Alley is somewhat compressed.

And yes, it is a shame that so few CD's have a dynamic range of 15 or more.

BTW the biggest selling album of all time has a DR of 15. It's a good recording.

 

[Thompsonuxb]

I think that if one were to take two examples of the same recording, let's say an original CD from the late 80's with a high DR value and a remastered, compressed version from recent years with a low DR value, the former will sound better on a good quality system (headphones in my case) and the latter will sound better in a car or on lesser quality speakers for reasons already given. I think it's pretty straightforward really. I've got some music that I can't listen to on headphones, it hurts my ears, but in the car it sounds fantastic.

Both flavours should be available - it can't be that hard.

I don't think that's true.

Quality CD's sound like quality CD's regardless.

I do alot of MixCD's for my personal use if they sound good on my set it sounds good in the car.

compressed stuff sounds compressed and not enjoyable. but then again I'm comparing with what I know well recorded stuff 'can' sound like.

Is English your native language? It's just that sometimes you don't seem to understand what I write. Maybe it's me.

?

What did I say that's so 'off'.

All I'm saying is to my ear regardless of format a well produced dynamic recording sounds better regardless of the quality of the equipment.

I'm not arguing or questioning your post just giving another pov.

 

[Thompsonuxb]

BigH for Kind of Blue my preference would be the original stereo vinyl version or the 2 track tape. For the tracks recorded with a sharp pitch I'd just adjust the speed down a bit when I played them.

I'd want to listen to the various CD versions, level matched, on my own system before deciding which version I liked the best.

There's a high chance that Eva Cassidy Live at Blues Alley is somewhat compressed.

And yes, it is a shame that so few CD's have a dynamic range of 15 or more.

?

BTW the biggest selling album of all time has a DR of 15. It's a good recording.

*WALL*

Honestly I want to ask why level match but fear if I do it'll cause an issue.

What's wrong with just comparing, see which sounds better at any level?

I'm genuinely curious. So please don't take this the wrong way.

 

[pauln]

Surely it's been explained to you enough times?

 

[pauln]

I think that if one were to take two examples of the same recording, let's say an original CD from the late 80's with a high DR value and a remastered, compressed version from recent years with a low DR value, the former will sound better on a good quality system (headphones in my case) and the latter will sound better in a car or on lesser quality speakers for reasons already given. I think it's pretty straightforward really. I've got some music that I can't listen to on headphones, it hurts my ears, but in the car it sounds fantastic.

Both flavours should be available - it can't be that hard.

I don't think that's true.

Quality CD's sound like quality CD's regardless.

I do alot of MixCD's for my personal use if they sound good on my set it sounds good in the car.

compressed stuff sounds compressed and not enjoyable. but then again I'm comparing with what I know well recorded stuff 'can' sound like.

Is English your native language? It's just that sometimes you don't seem to understand what I write. Maybe it's me.

?

What did I say that's so 'off'.

All I'm saying is to my ear regardless of format a well produced dynamic recording sounds better regardless of the quality of the equipment.

I'm not arguing or questioning your post just giving another pov.

Well, you said that what I said above wasn't true yet at the same time seem to be agreeing.

My point was that some compressed CD's sound painful (to me) on good kit but sound OK in the car whereas CD's with 'quiet bits' are tricky in a car or other areas with high background noise because you can't hear those passages.

 

[Thompsonuxb]

Surely it's been explained to you enough times?

Ok, I'll try and explain. Hopefully in a way that's not controversial or argumentative.

Context - comparing two recordings in a real world setting.

From my perspective most of us have what we would consider a sweet spot on our amps.

My amp has a reversed readout -100db to 0. Imo my amp sounds best between -50db and -40db.

If I am comparing recordings that is the range I'd use to compare.

Setting it to just -45db
Will not tell me how a recording will sound at the lowest levels I listen at nor the highest.

In the context of this thread ref the DR that's how I'd compare.

The 'explanations' given previously were under 'controlled' conditions between different equ.

That's all. That's why I asked.

 

[SteveR750]

Again I feel I have to chip in here.

Many of the complaints that I see in these debates (dynamic range, compression) is not about compression as such, not even over compression but about peak limiting.

We have all seen the traces of recordings where all the music peaks are pushed up hard to the 0dB line and this is, I think, the major problem. With the peaks 'cut off' in this manner we are virtually dealing with a series of square waves, which invariably bring a certain harshness to the recordings, a harshness that is particularly unpleasant on better equipment.

Sure, some compression techniques are pretty unpleasant too (side chain compression for example) but on most recordings it is the squaring off of overdriven peaks that causes the bigest problem.

I didn't think a brickwall limiter works like this, you're describing an analogue power supply rail clip?

 

[davedotco]

I didn't think a brickwall limiter works like this, you're describing an analogue power supply rail clip?

Very possibly, my understanding of things digital is pretty patchy.

I am refering to some of the traces that we often see on here, such as this.

This is a remastered version of Hotel California, heavily compressed with all the peaks driven up to exactly the same level and, in laymans terms, clipped.

This version sounds appalling and I do not think it is just down to lower dynamic range. There is a harshness that is analogous to analog clipping that is evident troughout and it is this kind of distortion that I think upsets a lot of people.

Anyone able to describe what is happening here in terms that I can understand?

 

[lindsayt]

That is a very good example Davedotco. It's the sort of recording where if you were to draw a volume vs time graph in your head as you listened, it would come out similar to the graph you've shown. Which makes for a frustrating listening exerience, as the original musical performance deserves a lot more room to breath.

Thompsonuxb, there's a natural tendency when doing A/B demos for the version played at the louder volume to sound more impressive. If you have a CD version that has 3 db less DR, as calculated by the DR database algorithm, chances are it will have a higher average recording level. So play the two versions, one after the other without touching the volume control and the more compressed version is likely to play back louder and therefore may give the initial impression of sounding more impressive.

This leads to the conundrum of how do you level match the two versions? This will always be a bit of a compromise. If you level match according to the loudest peaks, this will give a slight advantage to the more compressed version - in terms of "impressiveness". If you level match according to the quietest bits of the music, this will give an advantage to the less compressed version, as the transient peaks will come out louder. If you level match according to the average levels, this is fairer, but you'd then be looking at feeding your sound pressure meter (or voltmeter across the speaker terminals) into a laptop, which would then calculate the average volume.

 

[SteveR750]

That is a very good example Davedotco. It's the sort of recording where if you were to draw a volume vs time graph in your head as you listened, it would come out similar to the graph you've shown. Which makes for a frustrating listening exerience, as the original musical performance deserves a lot more room to breath.

Thompsonuxb, there's a natural tendency when doing A/B demos for the version played at the louder volume to sound more impressive. If you have a CD version that has 3 db less DR, as calculated by the DR database algorithm, chances are it will have a higher average recording level. So play the two versions, one after the other without touching the volume control and the more compressed version is likely to play back louder and therefore may give the initial impression of sounding more impressive.

This leads to the conundrum of how do you level match the two versions? This will always be a bit of a compromise. If you level match according to the loudest peaks, this will give a slight advantage to the more compressed version - in terms of "impressiveness". If you level match according to the quietest bits of the music, this will give an advantage to the less compressed version, as the transient peaks will come out louder. If you level match according to the average levels, this is fairer, but you'd then be looking at feeding your sound pressure meter (or voltmeter across the speaker terminals) into a laptop, which would then calculate the average volume.

It's not straighforward is it! I wonder how Spotify level matches it's files, as they all *seem* to be the same, I don't feel the need to keep changing the volume setting, whereas when I'm listenieng to my CD collection off J River with the volume levelling off, they are all over the place.

 

[davedotco]

What we all need is this........

 

[SteveR750]

What we all need is this........

Careful! You'll have the ear police calling.

 

[Glacialpath]

Napalm Death's 1990 release of Harmony Corruption has a DR of 14. Which is pretty good.

Compare this against their 2009 Time Waits for No Slave album with its DR of 5 to see which you think sounds better / more dynamic.

Harmony Corruption sounds better to me. The quality of the recording may not be as good a Time Waits for No Slave but from an overall listening and understanding of what is being played point of view then Harmony Corruption is better.

Enemy of the Music Business and Order Of the Leach were over limited and sound flappy.

The code is Red and all ND albums since then sound much clearer. A little too crisp maybe but that's the way they have been mixed and this whole flatlining everything has been honed a bit so things sound less flappy.

 

[TrevC]

That is a very good example Davedotco. It's the sort of recording where if you were to draw a volume vs time graph in your head as you listened, it would come out similar to the graph you've shown. Which makes for a frustrating listening exerience, as the original musical performance deserves a lot more room to breath.

Thompsonuxb, there's a natural tendency when doing A/B demos for the version played at the louder volume to sound more impressive. If you have a CD version that has 3 db less DR, as calculated by the DR database algorithm, chances are it will have a higher average recording level. So play the two versions, one after the other without touching the volume control and the more compressed version is likely to play back louder and therefore may give the initial impression of sounding more impressive.

This leads to the conundrum of how do you level match the two versions? This will always be a bit of a compromise. If you level match according to the loudest peaks, this will give a slight advantage to the more compressed version - in terms of "impressiveness". If you level match according to the quietest bits of the music, this will give an advantage to the less compressed version, as the transient peaks will come out louder. If you level match according to the average levels, this is fairer, but you'd then be looking at feeding your sound pressure meter (or voltmeter across the speaker terminals) into a laptop, which would then calculate the average volume.

It's not straighforward is it! I wonder how Spotify level matches it's files, as they all *seem* to be the same, I don't feel the need to keep changing the volume setting, whereas when I'm listenieng to my CD collection off J River with the volume levelling off, they are all over the place.

You can turn off the auto volume feature in Spotify, but it works better than most so I just leave it on.

 

[SteveR750]

Where is it? I haven't noticed it in the settings.... :O