Questions from a Roon newbie -

[klinemj]

May 7, 2018 16:53:33 GMT -5 audiosyndrome said:

May 6, 2018 19:50:57 GMT -5 klinemj said:

In my early days of ripping, I compared ripped versions of CD to the originals. I played the ripped CD via jRiver on a cheapo Dell 660s via USB to my DC-1. I played the original CD via an Emotiva CD player I used to have (ERC-2, I think) to the DC-1. I tried several selections and found no difference. The PC I used to rip the disks was a Powerspec (built by MicroCenter) with a run of the mill disk drive.

My feeling is that unless something goes wrong in the ripping, the ripped copy and the CD should sound the same. If you want - send me some example files of your amigo's rips vs. yours on a flash drive. Don't tell me which is which and I will happily listen to see if I hear a difference.

Mark

It may be possible for the ripped copy to sound better than the original CD. Ripping with error detection ON, results in an easier load for the playback system. A CD with many errors, can be thousands, requires that the CD player work much harder as it has to correct the errors on the fly. So maybe a difference. I think the majority agrees that the rips sound better.

Russ

While it may be "possible", I've never heard a rip that sounded better. And, I've never heard anyone claim that until you mentioned it...so, I think your final sentence is a stretch. I would say that most people I know who have ripped CD's and use that as a source find it very convenient. But, sound better? Not in my experience.

That said, I am looking forward to the test I proposed to Boom.

Mark

[Casey Leedom]

Dudes, just freaking compare the Digital Data. Anything else is just boring and ... not we'll thought out ...

Now if you want to argue reconstruction filters., how best to clock incoming packets data out to a bit-stream sink, etc., sure, then we have something to talk about. Or, if you want to talk about the best ways of doing Analog-to-Digital Conversions with the commensurate [and required] low-pass filters ... sure, that's interesting too. But once it's Digital Data, just stop worrying.

Casey

[audiosyndrome]

May 7, 2018 20:02:46 GMT -5 klinemj said:

May 7, 2018 16:53:33 GMT -5 audiosyndrome said:

It may be possible for the ripped copy to sound better than the original CD. Ripping with error detection ON, results in an easier load for the playback system. A CD with many errors, can be thousands, requires that the CD player work much harder as it has to correct the errors on the fly. So maybe a difference. I think the majority agrees that the rips sound better.

Russ

While it may be "possible", I've never heard a rip that sounded better. And, I've never heard anyone claim that until you mentioned it...so, I think your final sentence is a stretch. I would say that most people I know who have ripped CD's and use that as a source find it very convenient. But, sound better? Not in my experience.

That said, I am looking forward to the test I proposed to Boom.

Mark

My experience the exact opposite. General agreement amongst almost 200 NYC / LI audio club members; rips sound better.

Russ

[KeithL]

To be honest, I am very doubtful of the veracity of that result.
Unless one of the devices in the playback chain is functioning sub-optimally, there is no technical reason why they should sound different.
Assuming the rip is perfect, then the actual digital audio data in the ripped file must be identical to the data on the disc.
The only area that leaves for a difference to originate would be jitter; and, assuming proper design and choice of interface, jitter should not be a factor.

The suggestion that one or the other format might have more data errors - which I have heard claimed - is specious:
- a proper rip will have been verified to be bit-perfect to the original data
- after being processed by the two-levels of hardware error correction that are part of the Red Book playback standard, the data from a CD should also be bit-perfect
- a CD that is so badly flawed that the CD player cannot perform a bit-perfect error correction will usually also fail to rip successfully.... and is a pretty rare thing.

I would find it interesting to learn how the testing was performed, and the test protocols used. As someone who has designed many product comparison tests, for use on a wide variety of different systems, it occurs to me that it would actually be quite difficult to devise a test procedure that would adequately control for all the necessary variables in order to produce a useful result. At a minimum, both the ripped file and the CD would have to be played on the same device or computer, via the USB output, via the same DAC, and the DAC would have to be a modern model with an asynch USB input. Anything less stringent and you're going to end up comparing the sound of different DACs, or the sound of different types of inputs on the same DAC, all of which may introduce significant variations.

I would then want to repeat the test with multiple different player devices altogether - to rule out the possibility that a given player might, for example, introduce more jitter to files than CDs - or the opposite - due to differences in processing requirements. It isn't unusual for a given individual player device to handle one file format or another differently due to differences in processing requirements - but it tends to vary from device to device. It would also be interesting to know whether those who claim to hear a difference report that the difference "persists" after writing the rip to a CD-R disc, which is then played on the original player, or if the new copy is heard as sounding identical to the original disc (since we could then confirm that the bits are truly identical between the two discs).

Even assuming we're talking about informal tests, or even anecdotal evidence, I would still be interested to know the details of the conditions involved.... so as to recognize obvious flaws in the way the comparisons were performed - or acknowledge the lack of obvious flaws in the process.

[NOTE: I'm assuming that a ripped file was being compared to a physical disc being played "live". This is the case often stated of "comparing a file played off a computer to a CD played on a disc player". Of course, if we're talking about two different files, then they either are or are not identical. Files contain ONLY data; files do not and can not contain or record jitter; although jitter somewhere in the signal chain could conceivably corrupt the data before it's stored in the file, or affect the conversion process when it's converted back into analog. Remember that digital data is simply s stream of numbers, and a clock becomes involved at the point where those numbers are converted to and from analog audio; there is no additional mysterious factor involved. If excessive jitter, or uncorrected errors, resulted in corruption of the data as the file was being ripped or recorded, then the result would be clearly seen as a difference in the data - so the file would be different. The only way in which jitter could affect the signal audibly, assuming the files were in fact identical, would be if it was present at the exact point in the process where the D/A conversion occurs. Remember that "jitter" is not "a thing"; jitter is simply a term used to describe relatively short term errors in the clock... (when thinking about it, substitute "clock errors" for "jitter" in your internal dialog). However, because clocks are actually difficult to measure that accurately, and you don't actually listen to the clock anyway, we often tend to conflate the distortion in the audio signal that is caused by jitter being present during the D/A conversion process as being jitter; in fact, that is more properly "jitter sidebands", or, even more properly "distortion sidebands caused by jitter".]

May 8, 2018 17:18:44 GMT -5 audiosyndrome said:

May 7, 2018 20:02:46 GMT -5 klinemj said:

While it may be "possible", I've never heard a rip that sounded better. And, I've never heard anyone claim that until you mentioned it...so, I think your final sentence is a stretch. I would say that most people I know who have ripped CD's and use that as a source find it very convenient. But, sound better? Not in my experience.

That said, I am looking forward to the test I proposed to Boom.

Mark

My experience the exact opposite. General agreement amongst almost 200 NYC / LI audio club members; rips sound better.

Russ

[Boomzilla]

This'll be an empirical slam-dunk. Russel will rip some tracks from a CD to a USB stick, I'll rip the same tracks from the same CD to the same stick. Either the files will be identical or they won't. (I'm suspecting they won't).

If the files don't match, then no need to move forward. We've found the difference. One rip or the other isn't perfect.

If the files DO match, then it's off to klinemj for a blind review. If he identifies one rip or another as "better sounding" or even "different" without knowing which is which, then Houston, we have a problem.

If his findings are random, then all is as we expect.

End of discussion.

[DYohn]

Back to Roon: did you get it working properly? Did you like it?

[Boomzilla]

May 9, 2018 9:07:37 GMT -5 DYohn said:

Back to Roon: did you get it working properly? Did you like it?

Hi David!

Yes, I got it working. Yes, it streamed PCM over Ethernet without dropping down to MP3-320. Yes, I really liked the way it kept playing similar music to what I'd first selected. Yes, I liked all the information supplied about the artists.

No, I didn't like the way it lost half of my carefully-collected cover art. No, I didn't like the way it failed to import about 20% of my existing music library.

And, as the old saying goes, one AH-S**T wipes out a long list of "ATTA-BOYS."

So, taking into account the fact that I already own and am familiar with jRiver (despite its faults), I may stick with what I've got.

[Casey Leedom]

Boomzilla,

There are ways to tell Roon that you want your own tags to take precedence over their tags including cover art. Not sure how since I haven't tried it yet.

On the 20% of you music collection which hasn't been ingested yet, it might still be working on it. The scanning process take a very long time. I have a Ryzen 1500X system with super fast memory, NVMe root drive, 7200RPM music volume, etc. — i.e., not a slow machine. — and it took a very long time to scan my 4+TB of FLAC/DSDs (~6000 albums, ~100,000 songs). It was back about 5 months ago now so I don't remember the exact time, but I think it was on the order of a day even after I told it to go all out and use 8 CPUs ...

Casey

[Loop 7]

May 9, 2018 11:18:22 GMT -5 Boomzilla said:

No, I didn't like the way it lost half of my carefully-collected cover art. No, I didn't like the way it failed to import about 20% of my existing music library.

Is your cover art in the form of "folder.jpg" files in each album folder or are they embedded in the metadata?

Try this. Might be best to do this on the Macbook vs on a tablet.

On an album page, find the little menu drop down and click "Edit"




Select the "Edit Album" tab, scroll down and you should see the cover art that Roon selected (often based on pixel dimensions) and hopefully any cover art you curated or added to the album during prep.

[klinemj]

Boomzilla

On the ripped files - don't indicate whose they are...just tag them as different names. I'd rather not know which is which.

I'll listen first from my NAS via jRiver to my LH Labs "Geek" to my Maggies via my tube amp and then via my nCores. Then, I'll listen from my NAS via jRiver to my DC-1 to my Senn's via my BasX A100.

Mark

[Soup]

Would love to get in on this test. Can u pm me the files too?

I will run them thru microRendu to Chord 2Qute to Anthem mrx 510 (no processing mode).

Thanks

[klinemj]

May 9, 2018 15:03:43 GMT -5 Soup said:

Would love to get in on this test. Can u pm me the files too?

I will run them thru microRendu to Chord 2Qute to Anthem mrx 510 (no processing mode).

Thanks

After I do it, I could send the flash drive to you (save Boom from getting a bunch of flash drives...)

Mark

[Casey Leedom]

These aren't likely to be big files. Maybe a few dozen MB each. You could just put them up on an ftp server if you don't want to use email attachments.

Casey

[Boomzilla]

Missing the point, Casey - Once recopied, the "magic pixie dust" that's supposed to make these files "special" will be gone. The whole thrust of this claim (the one that you, completely and logically, reject) is that "bits aren't just bits."

And Mark - this means that you CAN'T copy the files to your NAS. You'll need to plug the USB stick into a destination and listen to the files DIRECTLY from the USB stick. A player is ALL you get to use between the USB stick & the DAC (and the simpler the player, the better).

And of course, the files will NOT be identified except by number. Two numbers will play the same cuts (one of Russel's and one of mine), but the order will be random. So, for example, if the same song is to be cuts 1 & 2. You can say:

1 is better
2 is better
No difference
There's a difference, but neither is superior

And I'll check before sending the USB stick to make sure the files ARE digitally identical. If they aren't, then either Russel or I are making imperfect copies, and this "experiment" need go no further.

[klinemj]

May 9, 2018 16:24:44 GMT -5 Boomzilla said:

Missing the point, Casey - Once recopied, the "magic pixie dust" that's supposed to make these files "special" will be gone. The whole thrust of this claim (the one that you, completely and logically, reject) is that "bits aren't just bits."

And Mark - this means that you CAN'T copy the files to your NAS. You'll need to plug the USB stick into a destination and listen to the files DIRECTLY from the USB stick. A player is ALL you get to use between the USB stick & the DAC (and the simpler the player, the better).

And of course, the files will NOT be identified except by number. Two numbers will play the same cuts (one of Russel's and one of mine), but the order will be random. So, for example, if the same song is to be cuts 1 & 2. You can say:

1 is better
2 is better
No difference
There's a difference, but neither is superior

And I'll check before sending the USB stick to make sure the files ARE digitally identical. If they aren't, then either Russel or I are making imperfect copies, and this "experiment" need go no further.

I have to admit - this part about the "magic pixie dust" if re-copied just doesn't make any sense to me. Can you elaborate on what would get lost? I could plug the stick into my headphone PC and output USB to my DC-1, I guess. But, I really don't understand what would get lost in transferring a file off a USB flash drive to a NAS.

Mark

[Casey Leedom]

May 9, 2018 16:24:44 GMT -5 Boomzilla said:

Missing the point, Casey - Once recopied, the "magic pixie dust" that's supposed to make these files "special" will be gone. The whole thrust of this claim (the one that you, completely and logically, reject) is that "bits aren't just bits." ...

Yeah, if that's the "question", then I'll just bow out of this one.

Casey

[garbulky]

May 9, 2018 16:24:44 GMT -5 Boomzilla said:

Missing the point, Casey - Once recopied, the "magic pixie dust" that's supposed to make these files "special" will be gone. The whole thrust of this claim (the one that you, completely and logically, reject) is that "bits aren't just bits."

And Mark - this means that you CAN'T copy the files to your NAS. You'll need to plug the USB stick into a destination and listen to the files DIRECTLY from the USB stick. A player is ALL you get to use between the USB stick & the DAC (and the simpler the player, the better).

And of course, the files will NOT be identified except by number. Two numbers will play the same cuts (one of Russel's and one of mine), but the order will be random. So, for example, if the same song is to be cuts 1 & 2. You can say:

1 is better
2 is better
No difference
There's a difference, but neither is superior

And I'll check before sending the USB stick to make sure the files ARE digitally identical. If they aren't, then either Russel or I are making imperfect copies, and this "experiment" need go no further.

This is a tree that there's no point in climbing.

There's a fundamental difference between CD rips (a copy) and say playing a digital file to a DAC (a stream).

Sending (streaming) a digital file to a DAC is prone to jitter. Sometimes things come slightly faster or slower than they are supposed to appear. A CD rip doesn't have this issue. There is an error correction mechanism. It's not about the timing anymore. If the bits on the hard drive aren't EXACTLY what's on the CD, then the system is let known that the two are not the same.

It's like they mailed a letter with a $100 in along with a note saying that "in this letter there will be exactly one $100 bill and this note". If the letter you opened didn't have those two items then you know you didn't get the identical letter that was sent.

That's why when you copy business documents, you aren't worried if you are suddenly going to lose a page out of your documents.

[monkumonku]

May 9, 2018 17:56:53 GMT -5 garbulky said:

May 9, 2018 16:24:44 GMT -5 Boomzilla said:

Missing the point, Casey - Once recopied, the "magic pixie dust" that's supposed to make these files "special" will be gone. The whole thrust of this claim (the one that you, completely and logically, reject) is that "bits aren't just bits."

And Mark - this means that you CAN'T copy the files to your NAS. You'll need to plug the USB stick into a destination and listen to the files DIRECTLY from the USB stick. A player is ALL you get to use between the USB stick & the DAC (and the simpler the player, the better).

And of course, the files will NOT be identified except by number. Two numbers will play the same cuts (one of Russel's and one of mine), but the order will be random. So, for example, if the same song is to be cuts 1 & 2. You can say:

1 is better
2 is better
No difference
There's a difference, but neither is superior

And I'll check before sending the USB stick to make sure the files ARE digitally identical. If they aren't, then either Russel or I are making imperfect copies, and this "experiment" need go no further.

This is a tree that there's no point in climbing.

There's a fundamental difference between CD rips (a copy) and say playing a digital file to a DAC (a stream).

Sending (streaming) a digital file to a DAC is prone to jitter. Sometimes things come slightly faster or slower than they are supposed to appear. A CD rip doesn't have this issue. There is an error correction mechanism. It's not about the timing anymore. If the bits on the hard drive aren't EXACTLY what's on the CD, then the system is let known that the two are not the same.

It's like they mailed a letter with a $100 in along with a note saying that "in this letter there will be exactly one $100 bill and this note". If the letter you opened didn't have those two items then you know you didn't get the identical letter that was sent.

That's why when you copy business documents, you aren't worried if you are suddenly going to lose a page out of your documents.

But if jitter is an issue relating to the conversion of digital to analog (like through a DAC), then why would a CD rip be different than a digital file? They both are digital content going to a DAC to be converted to analog, aren't they?

[garbulky]

May 9, 2018 18:12:08 GMT -5 monkumonku said:

May 9, 2018 17:56:53 GMT -5 garbulky said:

This is a tree that there's no point in climbing.

There's a fundamental difference between CD rips (a copy) and say playing a digital file to a DAC (a stream).

Sending (streaming) a digital file to a DAC is prone to jitter. Sometimes things come slightly faster or slower than they are supposed to appear. A CD rip doesn't have this issue. There is an error correction mechanism. It's not about the timing anymore. If the bits on the hard drive aren't EXACTLY what's on the CD, then the system is let known that the two are not the same.

It's like they mailed a letter with a $100 in along with a note saying that "in this letter there will be exactly one $100 bill and this note". If the letter you opened didn't have those two items then you know you didn't get the identical letter that was sent.

That's why when you copy business documents, you aren't worried if you are suddenly going to lose a page out of your documents.

But if jitter is an issue relating to the conversion of digital to analog (like through a DAC), then why would a CD rip be different than a digital file? They both are digital content going to a DAC to be converted to analog, aren't they?

The difference is that one is streaming in real time. As in what you get in a digital stream you have to play and you have to play it in the exactly right time. You will play it BEFORE the rest of the CD or the song gets there. And the protocols used to do this are very different from copying digital data. So here you can have jitter.

The other is related to copying digital data. You don't have to get it at exactly the right time. You don't have to play it right away before the rest of the cd or track gets there. Here the entire CD or entire track can be ripped and then compared to see if you have the correct data.

So in streaming you have a line of a painting. And more lines fill it up over the hour in hopefully the right order to create a painting of a CD. Now I'm simplifying. There are ways to make sure that the "hopefully" is a lot more accurate than just "hopefully".

In an exact audio copy of a CD, you get all the lines with a note coming with every line saying some letter of the alphabet. At the end of the CD you take the letters that came with each line and put them one after the other and you see that they make up the word "compact disc". If the letters said "co5mpac di6sc" then you know that this is not a bit perfect rip.

[monkumonku]

May 9, 2018 19:11:50 GMT -5 garbulky said:

May 9, 2018 18:12:08 GMT -5 monkumonku said:

But if jitter is an issue relating to the conversion of digital to analog (like through a DAC), then why would a CD rip be different than a digital file? They both are digital content going to a DAC to be converted to analog, aren't they?

The difference is that one is streaming in real time. As in what you get in a digital stream you have to play and you have to play it in the exactly right time. You will play it BEFORE the rest of the CD or the song gets there. And the protocols used to do this are very different from copying digital data. So here you can have jitter.

The other is related to copying digital data. You don't have to get it at exactly the right time. You don't have to play it right away before the rest of the cd or track gets there. Here the entire CD or entire track can be ripped and then compared to see if you have the correct data.

So in streaming you have a line of a painting. And more lines fill it up over the hour in hopefully the right order to create a painting of a CD. Now I'm simplifying. There are ways to make sure that the "hopefully" is a lot more accurate than just "hopefully".

In an exact audio copy of a CD, you get all the lines with a note coming with every line saying some letter of the alphabet. At the end of the CD you take the letters that came with each line and put them one after the other and you see that they make up the word "compact disc". If the letters said "co5mpac di6sc" then you know that this is not a bit perfect rip.

I could be wrong but I don't think that has anything to do with jitter; that is accuracy of the signal but it is still up to the DAC as far as how that is handled, and the DAC is where the jitter takes place, not on the source file.