Article: The Best Version Of… Steely Dan’s Aja

[boxitracio]

Estimado, es el más completo y documentado relato de la realización de Aja. Lo felicito sinceramente. Sus articulos son lectura obligatoria para mí. Saludod cordiales

[M1chael]

I have MFSL Aja CD, it’s one of my better/best sounding digital discs. Not because it’s MFSL but because it was recorded well. It does have what you refer to as u shaped frequency response, but my loudspeakers are so good that it still sounds great. I’m waiting before I pull the trigger on the Hoffman mastering until you check out the new SACD by Analogue Productions coming soon.

I can think of many CD’s you could TBVO, but I was wondering why most of The Rolling Stones ABKCO SACD’s sound pretty good. But when you get to Sticky Fingers & Exile they sound so dirty. Thank Mike

[Josh Mound]

For those still following this, I just added a huge update (in length and importance) centered on the new Bernie Grundman mastering.

[CDJones]

On 10/23/2021 at 3:03 PM, Josh Mound said:

For everyone still following this column, I've added a new update that may completely change the story of Aja on CD.

As for my own listening habits, I have to admit that the MQA CD of Aja has pulled into the lead for me (though I never, ahem, "unfold" the CD to MQA).

 

Thanks, Josh, for a fascinating article.  So... the MQA CD of Aja bests them all, without MQA decoding?  I'd be interested to hear your impressions should you get the opportunity to hear it with proper decoding (I have stayed away from MQA, but It's intriguing).

[Josh Mound]

5 hours ago, CDJones said:

 

Thanks, Josh, for a fascinating article.  So... the MQA CD of Aja bests them all, without MQA decoding?  I'd be interested to hear your impressions should you get the opportunity to hear it with proper decoding (I have stayed away from MQA, but It's intriguing).

 

I've listened to it with and without the MQA decoding. I don't think it makes a meaningful difference, positive or negative.

[CDJones]

Thanks, Josh.

 

Looking forward to the next TBVO!

 

 

[firedog]

Okay, I recently bought the "new" hi-res download that came out on all the download sites about a month ago. 

Not clear to me from the article: is it one of the above masterings, or something else? 

[Josh Mound]

34 minutes ago, firedog said:

Okay, I recently bought the "new" hi-res download that came out on all the download sites about a month ago. 

Not clear to me from the article: is it one of the above masterings, or something else? 

 

That's what I'm referring to as the Grundman. The same mastering should be on the upcoming SACD, too.

[Kalpesh]

I'm a great admirer, even more so since I started my own TBVO, comparing digital and LPs versions I own and I have to admit I do it quite loosely, matching levels by ears, and giving up before I find an explanation to what "digital editing" means for Neil Young's Le Noise while my original LP trounces the 24/44.1* files. 

However, I wish you share the response curves of your gears because otherwise comments on tonality are much less to the point IMO than comments on source and used gears (at least implicitly it can be guessed that Brown and Grundman's remastering benefit from progress in ADC, cables, power supply etc).

Of course we know our systems and can make decisions for ourselves but watching the comparison of my digital and analog response curves, you could say : ah ah look the response curves, what he likes is downed HF*. Anyway, what is for sure is that a mastering's tonality can not be judged independently of room (or headphone) 's response**.

*For those who have read thus far, no I'm not promoting Vinyl. I quit buying vinyl in 2013 and I started my TBVO with the last, Woodkid's Golden age, and found (and it was the case back then, leading me to quit) the Limited Edition a caricature of vinyl with the suspicion that digital was deliberately "caricature of digital". And this has been true of many of the albums I compared so far (entering year 2007 now and in my last years in vinyl I bought only "popular", including trip hop electronica etc). I have no hires version for those years, did not find any listed, guess most were 44.1 digital recordings. Half of the time Digital wins my Go To, even against quite expensive LPs like Crystal Castles II (119 € on Discogs) ;  Half of the time, vinyl wins. And the LPs costing the most often beat their digital counterparts, such as Le Noise or Heligoland. Sound wise the latter (Massive Attack) is very good as digital too ( a rare exemple when I could be lured to pick one for the other) and costs 50 (fifty) times less on Discogs...

 

**For that comparison I use a response curve with downed HF for digital but most of the time, outside this comparison,  I use response tailored after Harman for Dirac +1 dB target. My eQ for analog is analog and 4 useful channels only : can't mett digital precision though it taught me broader Qs and fewer points beat absolute frequency precision, because of time domain I guess. The digital filtering matters too : with my May I find Sinc Medium PCM 32 fs to be the closest to my analog gear.

Harman_Full Range+1dB_under_80Hz.txt

[Josh Mound]

On 10/20/2023 at 11:43 AM, Kalpesh said:

I'm a great admirer, even more so since I started my own TBVO, comparing digital and LPs versions I own and I have to admit I do it quite loosely, matching levels by ears, and giving up before I find an explanation to what "digital editing" means for Neil Young's Le Noise while my original LP trounces the 24/44.1* files. 

However, I wish you share the response curves of your gears because otherwise comments on tonality are much less to the point IMO than comments on source and used gears (at least implicitly it can be guessed that Brown and Grundman's remastering benefit from progress in ADC, cables, power supply etc).

Of course we know our systems and can make decisions for ourselves but watching the comparison of my digital and analog response curves, you could say : ah ah look the response curves, what he likes is downed HF*. Anyway, what is for sure is that a mastering's tonality can not be judged independently of room (or headphone) 's response**.

*For those who have read thus far, no I'm not promoting Vinyl. I quit buying vinyl in 2013 and I started my TBVO with the last, Woodkid's Golden age, and found (and it was the case back then, leading me to quit) the Limited Edition a caricature of vinyl with the suspicion that digital was deliberately "caricature of digital". And this has been true of many of the albums I compared so far (entering year 2007 now and in my last years in vinyl I bought only "popular", including trip hop electronica etc). I have no hires version for those years, did not find any listed, guess most were 44.1 digital recordings. Half of the time Digital wins my Go To, even against quite expensive LPs like Crystal Castles II (119 € on Discogs) ;  Half of the time, vinyl wins. And the LPs costing the most often beat their digital counterparts, such as Le Noise or Heligoland. Sound wise the latter (Massive Attack) is very good as digital too ( a rare exemple when I could be lured to pick one for the other) and costs 50 (fifty) times less on Discogs...

 

**For that comparison I use a response curve with downed HF for digital but most of the time, outside this comparison,  I use response tailored after Harman for Dirac +1 dB target. My eQ for analog is analog and 4 useful channels only : can't mett digital precision though it taught me broader Qs and fewer points beat absolute frequency precision, because of time domain I guess. The digital filtering matters too : with my May I find Sinc Medium PCM 32 fs to be the closest to my analog gear.

Harman_Full Range+1dB_under_80Hz.txt 434 B · 0 downloads

Most of my critical listening is done on widely measured headphones like the Focal Utopia. But maybe I’ll post my speaker setup’s response sometime, too.

[Kalpesh]

1 hour ago, Josh Mound said:

Most of my critical listening is done on widely measured headphones like the Focal Utopia. But maybe I’ll post my speaker setup’s response sometime, too.

absolute best eQ curve or best fit with one's system ?? I think it's a legit question 

 

do you endorse the attached (says deficient bass + 3-5K dip)? do you use the proposed eQ ?https://www.audiosciencereview.com/forum/index.php?threads/focal-utopia-review-headphone.22103/ 

[Josh Mound]

16 hours ago, Kalpesh said:

absolute best eQ curve or best fit with one's system ?? I think it's a legit question 

 

do you endorse the attached (says deficient bass + 3-5K dip)? do you use the proposed eQ ?https://www.audiosciencereview.com/forum/index.php?threads/focal-utopia-review-headphone.22103/ 

I don’t agree with the Harman headphone or IEM targets.
 

With my TBVOs, I always explain in detail what I’m hearing in each mastering. People are free to listen themselves and come to different conclusions. 

[Kalpesh]

20 hours ago, Josh Mound said:

I don’t agree with the Harman headphone or IEM targets.
 

With my TBVOs, I always explain in detail what I’m hearing in each mastering. People are free to listen themselves and come to different conclusions. 

would be great if you would publish the delta between two eQs, maybe we could see emerge a pattern and deduce the response curve used by this or that mastering engineer. In the above exemples, maybe the mastering engineers heard the same tonality but as it's always as through their system's response curve they measure differently

[Josh Mound]

10 hours ago, Kalpesh said:

would be great if you would publish the delta between two eQs, maybe we could see emerge a pattern and deduce the response curve used by this or that mastering engineer. In the above exemples, maybe the mastering engineers heard the same tonality but as it's always as through their system's response curve they measure differently

 

That would be an interesting project, but we’d need a lot more info that to figure that out. I’d say it’s probably impossible. First, the EQ curves I show from Har-Bal are the “mid” frequency balances. There’s also the “side” balance, or you could look at the left and right channels individually. The mid is the best single measure, but it’s not the only one. Second, the quality of the transfer (the tape itself, the tape machine, the machine’s calibration, the ADC, etc.) will all affect the sound of the master. Third, we don’t know if/when a mastering engineer is using a different studio or updating their studio’s gear. Fourth, we don’t know if the sound of a certain mastering really reflects what that engineer felt was the best mastering, or whether it’s what the client preferred. Etc. Ultimately, the best way to figure out what an engineer is hearing is to find out about their studio setup (if you can).

 

It is possible, though, to use Har-Bal and various other mastering software to brine one mastering’s EQ into pretty close alignment with another mastering’s EQ. I’ve done that several times, including in my recent Cream TBVO, to try to separate the effect of EQ from the effect of the transfer and other factors. It’s a helpful exercise in certain circumstances. 

[Kalpesh]

On 10/24/2023 at 12:23 AM, Josh Mound said:

 

That would be an interesting project, but we’d need a lot more info that to figure that out. I’d say it’s probably impossible. First, the EQ curves I show from Har-Bal are the “mid” frequency balances. There’s also the “side” balance, or you could look at the left and right channels individually. The mid is the best single measure, but it’s not the only one. Second, the quality of the transfer (the tape itself, the tape machine, the machine’s calibration, the ADC, etc.) will all affect the sound of the master. Third, we don’t know if/when a mastering engineer is using a different studio or updating their studio’s gear. Fourth, we don’t know if the sound of a certain mastering really reflects what that engineer felt was the best mastering, or whether it’s what the client preferred. Etc. Ultimately, the best way to figure out what an engineer is hearing is to find out about their studio setup (if you can).

 

It is possible, though, to use Har-Bal and various other mastering software to brine one mastering’s EQ into pretty close alignment with another mastering’s EQ. I’ve done that several times, including in my recent Cream TBVO, to try to separate the effect of EQ from the effect of the transfer and other factors. It’s a helpful exercise in certain circumstances. 

Thank you 

if I was as savvy as you with those tools I think I'd start with a CD mastered by Bob Katz since he published his target curve.

 

Qualitatively the attached exemple of comparison you made is reminiscent of the comparison of 2 of my in room responses, one being after Harman default for Dirac, the second altering it @ -0.4 /2000/0.48 -2/5000/0.54 -3.5/20000/0.17, after @mitchco's published measurement in 

Quantitatively the spread between my 2 responses is too large compared to your exemple but differences in your exemple could be consistent with differences between closer targets, such as as Katz's and a variant ending/inspired (HF) by Harman's Trained Listeners.

 

More than a icing on the cake of your excellent TBVO It would rather be another project (I'm not up to the task !) : reverse engineer  main audiophile mastering engineers in room curves starting with the known Katz's

 

The biggest caveat of course is to assume that mastering engineers want to hear the same level of, say, HF, and that they just tune depending of their in room curve.