• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Dire Straits – On Every Street - Review, comparison between CDs, SACD MOFI, Vinyls. What’s wrong with the MOFI and Back To Black vinyl records?

Jean.Francois

Active Member
Joined
May 31, 2022
Messages
103
Likes
367
Hello,

On Every Street” is the sixth and final studio album by the band Dire Straits, released on September 9, 1991. This album was released six years after the previous album, “Brothers in Arms”. “On Every Street” contains 12 tracks and was produced digitally.
On Every Street -- Small .jpg


For this review, 8 versions were tested: Vinyl record Back To Black and MOFI; CD from 1991,1996,2000 and MOFI; SACD MOFI and Tidal HD.

In terms of dynamics, it's the Mofi version that has the best dynamics compared to the 1991, 1996, 2000 and Tidal CD versions, as shown by the waveform zoom on the song "On Every Street" and the digital curves.
waveform -- On Every Street -- CD 1991 vs 1996 vs SACD - small .jpg

Waveform: CD 1991, CD 1996 and SACD MOFI​

As far as the vinyl versions are concerned, one might wonder about the quality of vinyl realization today.

The graph below compares the spectra of Back To Black and MOFI vinyl records.
Spectrum -- On Every Street -- Vinyl (white) vs Vinyl MOFI (blue) - small.jpg

Spectrum Vinyl Back To Black - 2024 (white) vs Vinyl record Mofi - 2024 (blue)​

For the Back To Black vinyl record, the cutting was done by Bernie Grundman, and there is attenuation above 16 kHz (yellow zone), even though the music still contains information. We've been noticing a problem with Bernie Grundman's cutting for some time now, as was the case with the latest "Dark Side of The Moon". It's a real shame for today's productions, because in the past, the engravings were really of high quality, without these defects above 15 kHz.

On the MOFI vinyl record, the bandwidth rises well above 20 kHz (yellow arrow), with a present signal up to at least 30 kHz. However, in the red zone we see a signal that shouldn't exist, as the digital master cuts at 22 kHz. There's an aliasing of the spectrum, a phenomenon that shouldn't appear on a vinyl record and isn't present on other albums like "Brother In Arms", which also has a digital master. It's a digital flaw found on analog vinyl! You can find full details of this problem here.

To have a vinyl record with a high technical quality is not so simple as we can see.

To see the impact on vinyl records and compare all versions, you can listen to sample from the 8 versions HERE, and also find the measurements for all versions.

Enjoy listening,
Jean-François
 
Thanks for the post, very interesting!
I own the original CD and seeing the comparison graph confirms my feelings: the original seems different, to my ears better (perhaps also because it was the first one I listened to, so it became my reference) than the reissues. . '96 seems to be "framed"
 
Last edited:
Had a look at https://magicvinyldigital.net/2023/01/05/dynamics-of-the-media-and-music/

Why do we often find 96 dB for 16 bits and 144 dB for 24 bits? The calculation that is made is a shortcut, in fact, we take 2 to the power of 16 (2^16) and we get 16777216 (that is 20xlog(16777216) = 96 dB), but in reality, the signal is alternating, that is to say that it goes from a value -Vmax to +Vmax with a minimum at 0, to encode the fact that the signal is positive or negative, 1 bit is needed, so there are only 15 bits left to indicate the value which can vary from 0 to VMax, with a VMax worth 2^15 or 32768 which corresponds to 20xlog(32768 ) = 90 dB for 16 bits. For the 24 bits, it is the same principle, we have 23 bits for Vmax, which gives 138 dB for 24 bits.

Might it be that we due to the sign bit have +32768 positive values and -32768 negative values totaling 16777216 values?
minimum at 0
We can't have negative values? Really?
 
Last edited:
Hello,

On Every Street” is the sixth and final studio album by the band Dire Straits, released on September 9, 1991. This album was released six years after the previous album, “Brothers in Arms”. “On Every Street” contains 12 tracks and was produced digitally.
View attachment 353975

For this review, 8 versions were tested: Vinyl record Back To Black and MOFI; CD from 1991,1996,2000 and MOFI; SACD MOFI and Tidal HD.

In terms of dynamics, it's the Mofi version that has the best dynamics compared to the 1991, 1996, 2000 and Tidal CD versions, as shown by the waveform zoom on the song "On Every Street" and the digital curves.
View attachment 353976
Waveform: CD 1991, CD 1996 and SACD MOFI​

As far as the vinyl versions are concerned, one might wonder about the quality of vinyl realization today.

The graph below compares the spectra of Back To Black and MOFI vinyl records.
View attachment 353977
Spectrum Vinyl Back To Black - 2024 (white) vs Vinyl record Mofi - 2024 (blue)​

For the Back To Black vinyl record, the cutting was done by Bernie Grundman, and there is attenuation above 16 kHz (yellow zone), even though the music still contains information. We've been noticing a problem with Bernie Grundman's cutting for some time now, as was the case with the latest "Dark Side of The Moon". It's a real shame for today's productions, because in the past, the engravings were really of high quality, without these defects above 15 kHz.

On the MOFI vinyl record, the bandwidth rises well above 20 kHz (yellow arrow), with a present signal up to at least 30 kHz. However, in the red zone we see a signal that shouldn't exist, as the digital master cuts at 22 kHz. There's an aliasing of the spectrum, a phenomenon that shouldn't appear on a vinyl record and isn't present on other albums like "Brother In Arms", which also has a digital master. It's a digital flaw found on analog vinyl! You can find full details of this problem here.

To have a vinyl record with a high technical quality is not so simple as we can see.

To see the impact on vinyl records and compare all versions, you can listen to sample from the 8 versions HERE, and also find the measurements for all versions.

Enjoy listening,
Jean-François

Very interesting - thanks!

A couple of observations and totally non-expert speculations:

1. Based on my admittedly casual look at the three waveforms, it would appear that the 1991 CD is more or less as dynamic as the MoFI, yes? The 1996 is of course brickwalled - yuk.

2. RE the high-frequency dip on the Grundman, could it have anything to do with impedance interactions or even just basic frequency response performance of one or more of the components in his mastering chain? He's known for using tube gear, and I wonder if the interaction of output and input impedances of two components could produce that high-frequency drop, or perhaps just the aging of the actual tubes in some of his gear? Alternatively, is it possible he's choosing to attenuate frequencies above 16kHz to aid with the cutting of lacquers and he feels it doesn't negatively impact the sound to do so? It's not like most adults can hear much of anything above 16kHz anyway, and Grundman himself I would include in that group.

3. RE the MoFi aliasing, my first guess would be that it's related in some way to the fact that they digitize their analogue transfers using DSD - but as a total non-expert I don't know if that hypothesis holds water.
 
Many thanks for this detailed analysis. I own both the 1991 and 1996 (SBM™) versions on CD, and much prefer the original release — my subjective impression is that it breathes more easily, while the 1996 master is noticeably more congested.
 
Had a look at https://magicvinyldigital.net/2023/01/05/dynamics-of-the-media-and-music/
Why do we often find 96 dB for 16 bits and 144 dB for 24 bits? The calculation that is made is a shortcut, in fact, we take 2 to the power of 16 (2^16) and we get 16777216 (that is 20xlog(16777216) = 96 dB), but in reality, the signal is alternating, that is to say that it goes from a value -Vmax to +Vmax with a minimum at 0, to encode the fact that the signal is positive or negative, 1 bit is needed, so there are only 15 bits left to indicate the value which can vary from 0 to VMax, with a VMax worth 2^15 or 32768 which corresponds to 20xlog(32768 ) = 90 dB for 16 bits. For the 24 bits, it is the same principle, we have 23 bits for Vmax, which gives 138 dB for 24 bits.
Might it be that we due to the sigh bit have +32768 positive values and -32768 negative values totaling 16777216 values?
1st, he mixed up 16 and 24 bit:
we take 2 to the power of 16 (2^16) and we get 16777216
No, we don't:
  • 2^16 = 65536
  • 2^24 = 16777216

2nd, re. the "shortcut":
so there are only 15 bits left to indicate the value which can vary from 0 to VMax, with a VMax worth 2^15 or 32768 which corresponds to 20xlog(32768 ) = 90 dB for 16 bits.
Yes, the difference between the sample value 32767 and 1 is 90 dB but that's not what the dynamic range is. The most correct definition, I think, is the ratio of "full scale signal" and "quantization error" variances (powers) as described in Quantization Noise 101: Where does SNR about 6N dB come from?. If however the "difference between the lowest and the highest level" is preferred (because maybe it's easier to understand), then "level" is a wrong concept, it should be "amplitude". The highest amplitude is 32767 (the signal swinging from -32767 to 32767) and the lowest amplitude is 0.5 (the signal swinging from 0 to 1). Then we get our 96 dB:
20 * log(32767/0.5) = 96.
 
Back
Top Bottom