High resolution audio(not impressed)

[TomSawyer]

Well those little 'data points' could mean life and death for some - simply because it could mean a smoother listening experience. I've said this before - hires is not just about dynamic range as 99% here proffer.

And thats exactlythe point (forgive the pun). People who don't believe the science can't help feeling that the more data points you have, the more accurately or smoothly the DAC will reconstruct the waveform and hence "believe" in hi-res. The science says, that 44.1kHz is all you need to describe perfectly and hence for the DAC to reconstruct perfectly a waveform. Remember at no point do you listen to the data, you only ever listen to a continuous analogue waveform so it's always "smooth".

Extracting two files from the same original, playing them through the same DAC, recording them with the same ADC and comparing them should allow us to put this one to bed. Perhaps I could also post the resultant files somewhere (maybe downsampled to 24/96 for size reasons) with randomly selected filenames and the forum could blind test and say which came by reconstructing the 24/96 and which from the 16/44?

 

[lindsayt]

Do you really think I need to listen to the 2007 / 2008 Foxtrot remasters (with a mediocre DR of 10) to know that they will sound noticeably more compressed than the earlier releases (with a better DR of 13)?

So all of a sudden measurements are more important to you than your ears. Have a listen, compare and get back to me.

This is one of those times when a key factor of the listening experience for me is backed up by a key measurement.

There's a game that anyone can play at home. Take 2 CD's / LP's. Play them back to back. Then guess which one has the greater DR, or if they have the same DR. Then look them up in the DR database and see if you're right. The 2 CD's / LP's can be the same album, or they can be different albums of the same artist. Or different artists.

I usually play this game with different albums of the same artist.

I have enough experience of playing this game to know that the compression on the 2007 /2008 version of Foxtrot would turn it from an enjoyable event into something more ho hum. From something that stimulates both sides of my brain to something that stimulates the left side of my brain more than the right.

 

[andyjm]

Well those little 'data points' could mean life and death for some - simply because it could mean a smoother listening experience. I've said this before - hires is not just about dynamic range as 99% here proffer.

And thats exactlythe point (forgive the pun). People who don't believe the science can't help feeling that the more data points you have, the more accurately or smoothly the DAC will reconstruct the waveform and hence "believe" in hi-res. The science says, that 44.1kHz is all you need to describe perfectly and hence for the DAC to reconstruct perfectly a waveform. Remember at no point do you listen to the data, you only ever listen to a continuous analogue waveform so it's always "smooth".

Extracting two files from the same original, playing them through the same DAC, recording them with the same ADC and comparing them should allow us to put this one to bed. Perhaps I could also post the resultant files somewhere (maybe downsampled to 24/96 for size reasons) with randomly selected filenames and the forum could blind test and say which came by reconstructing the 24/96 and which from the 16/44?

Tom, the link I mentioned before, 'Audio Diffmaker' will allow you to do exactly what you propose, if you record the output of the DAC a couple of times (16/44.1 and 24/96 for example) it will perform a null test between two separate runs. It does some fancy level and phase matching prior to doing the difference calc and saving a difference file. The software was designed originally to put to bed all the nonsense 'I changed my mains cable and a veil was lifted' claims, but strangely no mainstream HiFi magazine ever adopted it. One can only wonder why....

It should work just as well for different sample rates.

Foobar allows you to do random ABX testing if you want to do it in the privacy of your own home.

 

[lindsayt]

http://dr.loudness-war.info/album/list?artist=steely+dan&album=countdown

1974 vinyl and 1986 CD versions DR 13

1998 and 2014 versions DR 11

Maybe TrevC really does like the sound of excessive compression (distortion)?

I don't.

Audio compression isn't distortion, most classical music was compressed in the age of vinyl (it had to be with only around 45dB to play with) and almost all music played on the radio is compressed. I personally don't like the extreme compression used on Radio 2 because it sounds unnatural on some tracks. Forget the compression, listen to the music. Both of my examples have been remixed to perfection. Use your ears man.

Here's the online dictionary definition of "distortion":

warp, twist, contortion, bend, buckle, deformation, deformity, curve, curvature, malformation, disfigurement, crookedness, gnarl, knot

misrepresentation, perversion, twisting, falsification, misreporting, misstatement, manipulation

Using the online dictionary definition of the word, dynamic compression is very much distortion.

And this is where you and I very much differ. I regard dynamic compression as a pernicious form of distortion. I will notice it when it's there (especially when played back to back with a less compressed recording). I don't like it. Whereas you seem to be embracing it.

 

[TomSawyer]

Thanks Andy, I'll take a look. How useful it will be with regard to this particular debate on the forum will depend on what output can be extracted to post as evidence. A null test (or not), a set of comparible wave traces and a pair of sound files for the forum to blind test should allow the issue to be be concluded one way or the other.

I wonder how many would actually commit to the blind test?

 

[shadders]

Well those little 'data points' could mean life and death for some - simply because it could mean a smoother listening experience. I've said this before - hires is not just about dynamic range as 99% here proffer.

And thats exactlythe point (forgive the pun). People who don't believe the science can't help feeling that the more data points you have, the more accurately or smoothly the DAC will reconstruct the waveform and hence "believe" in hi-res. The science says, that 44.1kHz is all you need to describe perfectly and hence for the DAC to reconstruct perfectly a waveform. Remember at no point do you listen to the data, you only ever listen to a continuous analogue waveform so it's always "smooth".

Extracting two files from the same original, playing them through the same DAC, recording them with the same ADC and comparing them should allow us to put this one to bed. Perhaps I could also post the resultant files somewhere (maybe downsampled to 24/96 for size reasons) with randomly selected filenames and the forum could blind test and say which came by reconstructing the 24/96 and which from the 16/44?

Hi,

The theory states 44.1kHz is sufficient, but the actual implementation is not perfect. As such since we cannot produce a system that perfectly meets the relevant theories etc., we can improve by increasing the sample rate and bit depth.

Regards,

Shadders.

 

[TrevC]

Do you really think I need to listen to the 2007 / 2008 Foxtrot remasters (with a mediocre DR of 10) to know that they will sound noticeably more compressed than the earlier releases (with a better DR of 13)?

So all of a sudden measurements are more important to you than your ears. Have a listen, compare and get back to me.

This is one of those times when a key factor of the listening experience for me is backed up by a key measurement.

There's a game that anyone can play at home. Take 2 CD's / LP's. Play them back to back. Then guess which one has the greater DR, or if they have the same DR. Then look them up in the DR database and see if you're right. The 2 CD's / LP's can be the same album, or they can be different albums of the same artist. Or different artists.

I usually play this game with different albums of the same artist.

I have enough experience of playing this game to know that the compression on the 2007 /2008 version of Foxtrot would turn it from an enjoyable event into something more ho hum. From something that stimulates both sides of my brain to something that stimulates the left side of my brain more than the right.

Yet the reverse is true. The improvement is huge on the remaster. So much for your theory.

 

[lindsayt]

Yet the reverse is true. The improvement is huge on the remaster. So much for your theory.

You are very much mistaken if you think that there's a huge improvement in the dynamic range of Foxtrot by going for the DR 10 version instead of the DR 13 versions.

Anyone can go out and buy both versions of Foxtrot and make up their own minds.

Anyone can google "Foxtrot remaster" and see what they've been saying about the 2007 / 2008 version on other forums. Plenty of comments there about the compressed sound on this version.

 

[andyjm]

Yet the reverse is true. The improvement is huge on the remaster. So much for your theory.

You are very much mistaken if you think that there's a huge improvement in the dynamic range of Foxtrot by going for the DR 10 version instead of the DR 13 versions.

Anyone can go out and buy both versions of Foxtrot and make up their own minds.

Anyone can google "Foxtrot remaster" and see what they've been saying about the 2007 / 2008 version on other forums. Plenty of comments there about the compressed sound on this version.

The DR measure is not dynamic range. From memory, the algorithm compares peak signal level with long-run average signal level. I guess reasonable men could differ about the best way of measuring perceived dynamic range in a meaningful way, but DR is not the same as the engineering measurement of dynamic range.

It is quite possible that from time to time the algorithm throws up numbers which seem at variance with the experienced sound.

 

[TrevC]

Yet the reverse is true. The improvement is huge on the remaster. So much for your theory.

You are very much mistaken if you think that there's a huge improvement in the dynamic range of Foxtrot by going for the DR 10 version instead of the DR 13 versions.

Anyone can go out and buy both versions of Foxtrot and make up their own minds.

Anyone can google "Foxtrot remaster" and see what they've been saying about the 2007 / 2008 version on other forums. Plenty of comments there about the compressed sound on this version.

Unless you've listened and compared the two mixes you can't really have an opinion of what they sound like. I don't like the remix of Trick of the Tail or Selling England because they've made the vocals too prominent IMO, but the remix of Foxtrot is just what the album needed. I doubt anyone can hear a difference of 4dB of dynamic range.

Weird how you are happy to trust your ears every time except in this one instance. Why is that?

"the 2009 remixed remaster is an absolute must-own version of this album"

http://www.worldofgenesis.com/GenesisAlbumReviews.htm

http://www.jazzshelf.org/genesismixes07.html

 

[lindsayt]

TrevC, do you really expect me to go out and spend £5.50 on the 2008 CD, just to prove to myself that I was right all along?

And to buy something that I would never want to play again after the initial listening test?

And even if I did go out and buy it and then report the results of my listening tests, what's to stop you replying with a "Well you're wrong, Lindsay. The 2008 version sounds more dynamic to me." Or "You're just suffering from expectation bias, Lindsay."? In other words, do you give me your word of honour that you would be happy to respect the results of my listening tests?

 

[TrevC]

TrevC, do you really expect me to go out and spend £5.50 on the 2008 CD, just to prove to myself that I was right all along?

And to buy something that I would never want to play again after the initial listening test?

And even if I did go out and buy it and then report the results of my listening tests, what's to stop you replying with a "Well you're wrong, Lindsay. The 2008 version sounds more dynamic to me." Or "You're just suffering from expectation bias, Lindsay."? In other words, do you give me your word of honour that you would be happy to respect the results of my listening tests?

If you have the original CD and LP as I do and like the album it has to be worth it. Even if you don't want to spend any money it's on Spotify.

 

[Forever Young]

I've also had a play with high def audio. I can't hear a difference. It can be hard to know it is definitely hd audio if it's something you've downloaded, not always obvious in my experience. Some are flacs or whatever taken from cd or similar, while others turn out to be cd quality in surround. The other thing is that the improvements I expected in treble are not going to make any difference without spending a fortune on super tweeters. Even with super tweeters I doubt I could hear the difference, despite having very good hearing for my age, that of a young teen at 35. I listened to some in a showroom. I think if you're going to have a load of frequencies going through the system that you can't even hear, it's not going to improve the sound, just make it worse. It's different with some vinyl, which has a mostly uncompressed quality, especially in the bass. It's more felt than heard, just wish my speakers were up to playing down to around the 12hz mark also.