Best way to compare two releases of same work?

[miguelito]

I am a fan of Maria Joao Pires’s Chopin Nocturnes. I have a few versions of this (original CDs from 1996, SHM-CD from a few years ago, download from a “complete solo recordings” set). None of these files are “remasters” to my knowledge.

 

I am curious how the files compare and what is a good way to do this comparison. I have run these through MusicScope, and you can clearly see the newer versions have a little more high frequency information. 

 

What is a good way to compare these files? Can one do a diff somehow?

 

Thx!

[wwaldmanfan]

I don't want to hoard music, I only keep stuff on my playback drives that I actually expect to listen to. However, I have accumulated as many as five or 6 different masterings of some of my favorite albums. I try to pare it down to one preferred version, and relegate the alternates to backup storage. At this point in my life, I'm trying to simplify, not complicate things.

 

A cursory run through TT DR Offline Meter shows me the dynamic range and peak volume. Higher DR and lower volume generally sounds better, but not always. Then, I pick a track or three and run them through Audirvana consecutively to hear the differences. I'll confirm this on each of my setups, my main system and my desktop system with both speakers and headphones. I adjust the volume level by ear, for parity. It's all subjective, not too scientific, but who am I trying to please besides myself?

 

I have found, with some exceptions, that original releases have better SQ than remasters. Why pay $15-$20 for a hi-res download, when you can still find a better-sounding 1980's CD used for five bucks on Discogs, ebay, or Amazon? Money-grubbing record companies cannot fool these golden ears.

[Kal Rubinson]

1 hour ago, miguelito said:

I am a fan of Maria Joao Pires’s Chopin Nocturnes. I have a few versions of this (original CDs from 1996, SHM-CD from a few years ago, download from a “complete solo recordings” set). ...............................................

What is a good way to compare these files? 

I would just listen to them because I find it pretty easy to decide which ones I prefer.  Measurements cannot help me because my personal enjoyment is the only important criterion.  All the rest is for demonstrations.

[james45974]

10 minutes ago, Kal Rubinson said:

I would just listen to them because I find it pretty easy to decide which ones I prefer.  Measurements cannot help me because my personal enjoyment is the only important criterion.  All the rest is for demonstrations.

I find this hard to do.  I have tried to compare different masterings of the same recording and more often than not prefer the one that I am more familiar with.  I agree, measurements don't help me either.

[89reksal]

1 hour ago, miguelito said:

I have run these through MusicScope, and you can clearly see the newer versions have a little more high frequency information. 

 

For me, if you're using Musicscope, I find that the version with the highest "Crest" (Crest Factor) value will be the best sounding version. Loudness Range (LRA) can be a factor as well if Crest of different versions are very similar. 

[Musicophile]

I find different masterings to be hard to tell apart by measurements, if they are in the same format, eg. 16/44 redbook.

 

I use Musicscope to check if high-res albums are not up-sampled

 

Dynamic range is an indicator but not a reliable signal of a better mastering. 

 

So in the end, it's down to you and your ears.

 

By the way, please keep us posted on your preferred version of the Pires Nocturnes. 

 

I'm a big fan as well.

[kilroy]

You can run them through foobar's (or other) dynamic range plugin to see that aspect. Changes between DR can indicate a different mastering. EAC can give you overall levels referenced to 0 dB.

[miguelito]

1 hour ago, wwaldmanfan said:

I don't want to hoard music, I only keep stuff on my playback drives that I actually expect to listen to.

True but...

 

1- There are some masterings that actually sound subtly different

I prefer to have them both available. Brutal case: Nina Simone's "Little Girl Blue" - Jeezus! My fave is def the new Acoustic Sounds DSD (I ripped the SACD with a PS3). But the mix is so different that it is interesting to listen to them all (I have 6 versions last time I checked). 

 

2- The Pires recording is my favorite Chopin Nocturnes - and all versions are from the same digital master so very close - but not identical. 

In this case, DR is almost identical so it's not telling me much. I would like to nail down my fave and it will boil down to listening critically. I do not intend to keep the many versions around in this case.

 

 

[miguelito]

6 minutes ago, kilroy said:

You can run them through foobar's (or other) dynamic range plugin to see that aspect. Changes between DR can indicate a different mastering. EAC can give you overall levels referenced to 0 dB.

DR is about identical in all cases, but the files are not identical.

 

In terms of preference it will boil down to listening, I think.

 

I am still curious about doing a diff of some sort and listening to that diff. Just cuz I'm crazy that way...   (  @mansr ?....  )

 

[rando]

Luck favors the prepared mind.  All the prep work to sort out where differences exist, and in what manner, helps considerably.  If you keep them in a background frame when it comes time to sit down and listen. 

 

If you are asking how to listen.  That is a very personal answer to creating a situation that helps promote best results.  I would like to point out we as consumers are growing very close to having the ability to take these results and combine the most desirable circumstances into one file.  If not a subtly morphing version that expands on what we deem most enjoyable.  Food for thought when approaching decision making time.      

[mansr]

37 minutes ago, miguelito said:

DR is about identical in all cases, but the files are not identical.

 

In terms of preference it will boil down to listening, I think.

 

I am still curious about doing a diff of some sort and listening to that diff. Just cuz I'm crazy that way...   (  @mansr ?....  )

There are lots of ways the files might differ. Do you know anything about the provenance of the recording? If it was recorded to analogue tape, you could be looking at different digitisations. If it was a digital recording, or the same digitisation was used, there could still be minor edits.

 

I'd start by looking at a spectrogram of each version and checking for anything that stands out as different. Zooming in on the first 10 seconds or so may be helpful too. If you post Musicscope screenshots, perhaps someone will spot something.

[miguelito]

13 minutes ago, mansr said:

There are lots of ways the files might differ. Do you know anything about the provenance of the recording?

Thanks @mansr. Will post musicscope screenshots over the weekend. You can see in them that the two recordings look very similar with a bit more high frequency content in the newer ones. This is relatively easy to see bc solo piano doesn't have a lot of high frequency content.

 

My understanding is the recordings were done directly to digital around 1993 (CD released in Oct 1996, I bought it as soon as it came out). 

[miguelito]

44 minutes ago, rando said:

If you are asking how to listen.

I am not really asking that, but always keep an open mind in case you have a recommendation...

 

I personally look for piano tonality to be "right". I have way too many recordings where the piano just doesn't sound right. In solo piano I don't actually value that much soundstaging - frankly the live sound of a piano is rarely well defined in space. I do look for notes sounding effortless, with the appropriate decay, and lack of distortion. All hard things to get in a recording.

[miguelito]

One notable case: Jarrett's Koln concert, where I love how you can hear how fecked up that piano was... And yet he made it shine.

 

I think the MQA version of this album is superb, btw...

[rando]

58 minutes ago, miguelito said:

I think the MQA version of this album is superb, btw...

 

/permaban

 

[miguelito]

4 hours ago, miguelito said:

One notable case: Jarrett's Koln concert, where I love how you can hear how fecked up that piano was... And yet he made it shine.

 

I think the MQA version of this album is superb, btw...

I have many versions of this album, including a high res one... MQA is pretty good. Not better than the high res one though, in my estimation.

[Kal Rubinson]

7 hours ago, james45974 said:

I find this hard to do.  I have tried to compare different masterings of the same recording and more often than not prefer the one that I am more familiar with. 

Nothing wrong with that.  I tend to keep everything since file storage space is cheap but, after a while, my preferences take hold and I play them.

[miguelito]

Here go the musicscope selfies...

 

1996 CD:

 

SHM-CD:

 

From "Complete Solo Recordings" (Presto Classical):

[miguelito]

Ok... So I’ve compared Nocturnes 2, 8, 13 - some of my faves. I have listened to this album in its entirety easily 1,500 time over the last 20 years. Indeed. So I know this recording pretty darn well.

 

The versions compared are:

1- Original CD - purchased when it came out in Oct 1996, ripped ~4yrs ago with XLD

2- SHM-CD - purchased last year out of curiosity - in case it was better - ripped with XLD

3- Presto Classical download this month - “Complete Solo Recordings”

4- TIDAL version of the album

 

Long story short, the versions sound pretty much identical to me. 

[89reksal]

1 hour ago, miguelito said:

Long story short, the versions sound pretty much identical to me. 

From the Musicscope results you posted, it shows Crest is identical for all and Loudness Range is virtually identical. Based on that, I would expect them to sound identical. The only noticeable difference is the True Peak Level (TPL) for the Presto Classical version is a little higher than the other two. All else being equal, I would expect that version to be the "better" sounding one of the three, although it would be very minor. If I had to keep only one version, the Presto Classical version would be the one for me.

[miguelito]

59 minutes ago, lasker98 said:

If I had to keep only one version, the Presto Classical version would be the one for me.

Well, I don't have to keep one version, but yes, I agree.

[bmoregnr]

For sure the final arbiter Is with your ears and better yet if everything is level matched so as not to fool you with loudness. 

 

I find the graphs and numbers can reinforce a narrative for lack of a better word, but there is always so much more like transfer gear etc. you really have to go with what you hear as best. 

 

One thing that may help is EAC, or DR numbers with peak levels so you can compare multiple versions that you think may be using the same remastering.  Even then there is an argument about different CD pressings of the same remastering being different— something that is a rarer occurrence and is negated when playing by files. 

 

It sounds like you need to spend some time on the Steve Hoffman Music Forums as these preference comparisons are going on constantly; and although I don’t see quite as much classical music discussed I’d guess somebody can provide similar classical forums for the purpose.

[HardrockInMiniMac]

On 16.2.2018 at 3:05 PM, james45974 said:

I find this hard to do.  I have tried to compare different masterings of the same recording and more often than not prefer the one that I am more familiar with.  I agree, measurements don't help me either.

 

That is why studies have proven kids would prefer mp3 over original versions. 

[wwaldmanfan]

 

On 2/26/2018 at 4:33 AM, HardrockInMiniMac said:

That is why studies have proven kids would prefer mp3 over original versions. 

 

Do you mean they preferred the SQ of a modern compressed and limited remaster over the original, or the mp3 was preferred for convenience because it fits on their phone?

[HardrockInMiniMac]

Preferred SQ