[WGH]

I'm an IT Professional with over 30 years of experience...

Last year I wanted to add 4K video to my audio workstation.  My base system could already support it: 32 CPU cores, 128GB of RAM, and a RTX3070 video card.  I upgraded to a 42" 4K HDR monitor and put a Sonos soundbar on my desk.  I found the right software to get things set up, but making it all work involved a huge learning curve (even for an IT expert  such as myself), a lot of trial and error, and a lot of frustration...

I have a library of 250 4K UHD movies ripped to MKV files on a 18TB hard disk.  Is it something I would recommend...maybe.  But only to those who have a high tolerance for goal frustration and masochistic tendencies where computer hardware and software integration are concerned.

I think you have too much knowledge to build a video server although all those cores may be useful to rip Blu-ray into .mkv files.

These days a 12 year old could build a video server in an hour. Plug an Intel i7 11700 in an Asus Prime H570M-Plus. Add 16 GB ram and, just for fun, a M2 SSD for the OS and bingo, you got a video server. Windows 11 and the free VLC Video Player completes the package. The i7 11700 will play 4K - 4096x2160@60Hz over HDMI. NO graphics card needed! It will even do dual monitor so the computer can used for something else while watching a movie (DVI-D & HDMI).

Too cheap? The i7 has a 65W TDP so the leftover money can be spent on a fanless case.

The above example isn't an exercise in theory, the i7/Asus combo is what I use every evening to watched ripped movies in a 7.1.4 setup. Everything I throw at it plays flawlessly without a glitch or shudder in Dolby Atmos or DTS-HD MA when available.

[jsaliga]

I think you have too much knowledge to build a video server although all those cores may be useful to rip Blu-ray into .mkv files.

With all due respect, I don't think you know what you're talking about.  You don't know what workloads I run on this system, or why it has the number of cores and amount of memory it does.  So you're just talking through your hat.

These days a 12 year old could build a video server in an hour. Plug an Intel i7 11700 in an Asus Prime H570M-Plus. Add 16 GB ram and, just for fun, a M2 SSD for the OS and bingo, you got a video server. Windows 11 and the free VLC Video Player completes the package. The i7 11700 will play 4K - 4096x2160@60Hz over HDMI. NO graphics card needed! It will even do dual monitor so the computer can used for something else while watching a movie (DVI-D & HDMI).

You are wrong on the facts, and this is not just a mere difference of opinion.  You don't have enough information to make a statement like that.  You're assuming that what works for you will work for me.  You don't even know what my use case is, what hardware I have integrated and why, or even what software I am using and why.  What you spell out above would not work for me at all, and I really have no desire whatsoever to explain it to you.

I think I have said all I care to say in this thread.  This kind of shoot-from-the-hip back and forth is just not my thing.  So I'm done.

--Jerome

[MttBsh]

I have enjoyed AudioCircle for many years but often wonder why simple questions posed here can end up generating so much conflict. I applaud the many members who are willing to spend their time answering questions and helping newbies, it's a great place to learn more about our hobby. Why anyone feels the need to get personal or take an offensive tone is beyond me.     

[WGH]

????

Jerome,

You are reading a lot more into my reply than I intended, my apology if I offended.

My reply was to illustrate that if someone wanted a simple media/streaming computer and not rip Blu-ray disks then the expense and expertise needed to build a work station is not needed.

If you think my setup doesn't/can't work, I invite you to come over anytime you are close to Tucson, AZ, we'll watch a movie. I'm really a nice guy. You can sleep on my couch.

[jsaliga]

* Sigh *  I'm sure I will end up regretting this.  The discourse here has already gone off the rails.

If you think my setup doesn't/can't work...

Why even bother with this??  Honestly.  I already said what you are doing will not work for my specific use case.  Why not just accept that?  I'm happy it works for you.  I truly am.  Not acceptable at all for my situation.  Why press??

To Scott,

The whole point of the comment where I first brought this up was just to empathize a little bit with you buddy.  I was sincere when I said that power users and experts often gloss over the technical details and the knowledge required to pull off a specific implementation.  Well sure, everything is a piece of cake when you know what the hell you're doing.  You found dealing with the computer an aggravating experience, but implementing a nice vinyl setup has a learning curve all its own.  The difference for you is that you have mastered the vinyl learning curve so to you it is simple to deal with.  But put a raw beginner in front of a turntable and it would intimidate the hell out them.

It was not my intention for the thread to turn into the can of worms it has become.

Later,

--Jerome

[sunnydaze]

Why anyone feels the need to get personal or take an offensive tone is beyond me.   

Some folks are super touchy and defensive by nature.  Usually followed by pouty and petulant little proclamations of "I'm gonna take my ball and go home!".  I see it on here all the time....from certain members.   

[jsaliga]

Why anyone feels the need to get personal or take an offensive tone is beyond me.   

It's always amusing to see someone act surprised when no one wants to step into the pool for a swim after they have thrown a proverbial turd into it. 

--Jerome

[dlaloum]

Quote from: dlaloum on  3 Jul 2022, 05:04 am
1) well recorded digital is superior to Vinyl - well, maybe, maybe not
.... but Digital has NO wear.  But hard drives crash on a regular basis.

by all objective, measurable criteria, digital, in any of its better current iterations is superior to vinyl. (it may be debatable for the highest level of high speed tape ... but not for vinyl) - there are a bunch of issues on vinyl that can be ameliorated, but never totally removed, and which affect the sound - eg: tonearm mass/compliance resonance, pops-clicks, noise floor.
Some of the flaws can be temporarily ameliorated - eg: deep ultrasonic cleaning (remove many pops/clicks), use lubricating treatment (lower noise floor, reduce wear, increase life of stylus), proper compliance / arm mass matching and use of fluid damping (or magneto/electric equivalent)... but that amelioration is temporary (records have to be cleaned with each use!) - and wear is always ongoing, it is fundamental to the nature of vinyl (unless using something like the finial laser TT... no contact vinyl!)


As for Hard drive reliability - anyone who works in IT, and many who don't, are familiar with concepts such as redundant arrays of disks, mirroring and backup... - I use an array with dual redundancy - and I back it up to seperate offline disks periodically.

[dlaloum]

That is not entirely true.  CD Audio is brickwalled at ~ 22KHz, or the Nyquist frequency of its 44.1KHz sampling rate.  See the spectrum below from a CD track loaded into Adobe Audition and note all of the black space above 22KHz.



Vinyl records commonly have content that far exceeds the 22KHz limit of CD audio.  My phono cartridge/phono preamp combo can reproduce frequences out to about 65KHz.  Here is the spectrum of a recent vinyl recording I made using a 192KHz sampling rate in Adobe Audition:



Note the spectrum extends out to about 65KHz.  There may be more information on the vinyl, but if there is my cartridge/phono preamp cannot reproduce it.  This is why I record vinyl at 24bit/192KHz.  I want to capture the full frequency range of a vinyl record that my cart/phono pre can reproduce.

Naturally, it is fair to point out we can't hear frequencies above 20KHz.  However, I have seen some studies on human ultrasonic frequency perception.  I don't claim to understand everything in those studies.  But my feeling is why not just capture everything on the record and erase any doubt about the fidelity of my digital captures.  Since storage is fairly cheap these days I record everything at 24bit/192KHz and occasionally I do DXD 24bit/352.8KHz even though I think it is overkill for my vinyl set up.  I mostly do it to address potential IMD and only then on what I would consider a prized recording.

--Jerome

There are a number of assumptions there that bear questioning:

1) The fact that there is SIGNAL above 22kHz - does not imply that there is intentional recording above 22kHz - it could be various types of noise, distortion, etc...

2) Even if there is actual intentional recording above 22khz - is it in any way audible by the listener? - the AEC ran a test of Hypersonic audibility at one of their conferences, and I think it was circa 1/50 professional audio engineers who could with statisticaly significant results, show that they could tell anything was different between a recording with or without the hypersonics - so 49/50 people, who base their career on listening, cannot hear hypersonics!!

3) Few and far between are the styli which can accurately trace a signal above 22kHz - mostly line contact needle types...  and even if the needle itself can trace that signal, unless the cantilever resonance is well controlled to beyond the desired signal frequency, which requires ultra low effective needle mass - you will get distortion of the signal... (as an example - the Beryllium walled cantilever from the V15VMR - resonant frequency is 32kHz, Jico SAS - I have two examples of Jico SAS, one has resonant F around 14kHz the other around 16kHz - even though both use exotic boron cantilevers. - sadly that results in boost in Harmonic distortion in a bell curve peaking at those frequencies. - I also have a dynavector karat, with resonant F=50Khz ... which keeps things nice all the way to 20Khz - the resonance boost bell curve starts to kick off just above 20Khz on that one. (on the V15V the resonant bell curve affected down to around 16kHz, but the rise was controlled with the cartridge loading - keeping FR flat... but it will still show on THD measurements)

Basically - even if you assume that there is USEFUL signal above 20Khz (which is highly doubtful) - most cartridges, even very high end ones, won't trace it or reproduce it accurately.

Then there are the speakers - most speakers have very limited signal reproduction capabilities over 20Khz - it varies depending on the design of the tweeters, but most roll off quite quickly after 20khz. (just so happens mine don't, having usefull signal to beyond 35khz for what that's worth!)

So there is all the logic and reason speaking....

Having said all that, when archiving LP's to digital, I use cartridge/stylus combos that are capable of tracking/tracing out to 48Khz (that I have measured), with appropriate Styli and phono stage. - And I adjust and tune the setup using test records, to ensure as close as is possible to a flat frequency response (I strongly prefer MM cartridges for this reason, as the loading allow some balancing of the cantilever resonance, allowing a flat FR, whereas with MC's the raw response of the cantilever is exposed - often harshly!).

Also, I record using 96/24, which means I am recording out to 48kHz - so the resulting recording contains anything that was there in the signal up to 48kHz

Having said that - my Dirac Live Room/Speaker EQ (one of the biggest improvements in my home setup in the last 30 years!!!) - is limited to 24Khz as it's max resolution in 48k. (Some of the higher end versions us higher digital frequencies, and therefore have higher theoretical bandwidth) 

In the comparison of trade-offs - the improvements gained from that Room EQ, are obviously and immediately noticeable to all listeners - possible gains from the presence of ultrasonics, have never been found to be audible in my system, or by anyone I know. (did some experiments years ago with supertweeters.... not convinced)
 
I do my recording monitoring using headphones (ie bypassing the room eq, receiving full bandwidth) - but even then, I doubt they have much useful response over 20Khz (I have Stax ESL HP's as well as Beyer DT880's and a couple of others)


So yes, I am well aware of vinyl's ultrasonic capabilities (I use them sometimes to measure cantilever resonances.... although my digital recording and analysis is limited to 48khz)

And all the above is the reason I am not fussed about downsampling from 96/24 to 44/16 ... in the final mastering stage, the 16 bit resolution (96db dynamic range) is plenty for any recording I have come across, once you have removed the peaks caused by clicks/pops - but you desperately need every bit of the 24bit resolution in the initial recording to capture those clicks/pops with minimum distortion, so they can then be edited out cleanly. (and that is after my extreme, ultrasonic++ cleaning routines!)

Once digital recording gear becomes available that has true useful dynamic range higher than 24bit (144db) , I will consider an upgrade - but right now, even the best systems tend to max out at around 20bit DR (120db).... even an ADC capable of FULL 24bit DR is yet to be found! - the additional DR, would make recording vinyl cleanly a lot easier....

[dlaloum]

   What I do is upsample a recording multiple times, in this case vinyl. Re-record it and upsample again. The end result is a better recording than what was originally recorded.

How do you know that to be the case? are you sure it isn't simple confirmation bias? - or is this a subjective opinion?

[dlaloum]

I don't think that's a valid point.  At least it is not consistent with my experience as an IT professional, where our datacenter has thousands of hard disks and we might have one fail in a year.  But truth be sold we don't sit around waiting for stuff to fail and have an active refresh program for storage and servers...usually every 5 to 7 years.  Personally I have not had a hard disk fail in over 10 years.  I currently have about 32TB of storage on my home network.  Moreover, storage is dirt cheap these days.  A 12TB USB 3.0 hard disk will set you back a little over $200.  That is a massive amount of storage, even for someone such as myself who has about 5,500 albums in my digital library, mostly in 24bit/192KHz FLAC, DXD 24bit/352.8KHz FLAC, and DSD256.  I keep both an online and offline backup of my entire digital music library and the storage to support it cost me less than $350.

So the advantages of digital music can't reasonably be argued, it seems to me.  I'm a vinylphile too, and it has some advantages that digital lacks.  But it also has some weaknesses as well that do not plague digital music.  In short, both have their strengths and weakness.

--Jerome

Amen!

32Tb must be the sweet spot.... that's where my home server sits at...

On the other hand, I have had several hard drive failures over the last 30 years... but have never lost data, as they were always in redundant arrays.

The maths around error likelihood vs array failure & rebuild time starts to become concerning when you get to 32TB.... but dual redundancy provides a decent level of protection.... and then there are backups.

[dlaloum]

I don't wish to be insensitive or brusk but the "my stereo" playing videos has always been a pet peeve, They never struck me as a rational means of conveying what one's stereo sounds like. If the video contained clips with different components then, yes, there's a basis for comparing differences. Other than that.... 

FWIW, Vinyl sounds better to me on my system when listening closely. Maybe it's me; maybe it's the system; maybe it's both.

I digitize my records. Yes, I hear a difference between them. To me, the difference is clear, maybe even obvious, in A/B comparison. Some folks say maybe they hear the difference or they hear an insignificant difference. Some others say it sounds the same. The digital version is not horrible or unlistenable. Non-A/B is reasonably close--for some value of reasonably and close. Many of the records I listen to are historic live performances. The difference in sound quality between the better and worse recordings is far greater than the difference between my analog originals and my digital copies. Digital has the advantage when it comes to portability, user convenience and wear characteristics compared to vinyl. I play records when I want to sit back & focus only on the music. I listen to digital when allowing myself to simultaneously attend anything else.

I asked Gabe Wiener, uber-perfectionist, founder of Quintessential Sound and PGM Recordings & chairman of the New York Section of the AES why his productions were only available on CD. He said, "I can make great sounding LPs and CDs. Digital is easier."

Thanks for stirring up memories of my friend. As always, YMMV.
/mp

So you find the original vinyl sounds better than the recording you made from it?

Have you checked the digital to confirm that no point in the recording reaches beyond -1db?

Lots of DAC's start to misbehave in the last db before clipping at 0db, and once it clips... typically with a pop or click - it impacts the entire frequency range on not just the bit that peaked....

This is why I record well down, with peaks at circa -20db, so I have loads of room for the unexpected pop - once I have captured it intact, I can edit out the pop, and re-level the recording to bring it within a 16bit dynamic range (96db) - where for the initial recording I need to have 116db to provide for the 20db overhead - hence the need to record with a decent 24bit ADC... and the be careful with level setting through the chain...

I have not found my end result to be consistently differentiable from the original.

But my early digital recordings... before I learnt some of these "tricks" (and upgraded my ADC) - were on occasion - audibly different from the vinyl.

(P.S. Videos of people's stereo playing.. ?!?! makes no sense whatsoever - always found the idea ridiculous!)

[jsaliga]

There are a number of assumptions there that bear questioning

Forgive me for declining to address your many points, the majority of which are incorrect on the technical/scientific merits.  I'm just not willing to invest the time it would take me to write a well articulated and cogent rebuttal with reference links that you most likely would dismiss out of hand.  You are free to consider that a failing on my part, but, in short, there are much better and more productive uses of my time...which is in short enough supply as it is.  And you clearly have made up your mind on these points so any time I spend debating this with you will have just gone up in smoke.

So, in response to your very well thought out post I will say this:  I capture and play back at 24bit/192KHz because the majority of studio professionals and mastering engineers agree that it is transparent to the source.  And even if they didn't I don't really need to justify what I do...to anyone.

Cheers.

--Jerome

[jsaliga]

Lots of DAC's start to misbehave in the last db before clipping at 0db, and once it clips... typically with a pop or click - it impacts the entire frequency range on not just the bit that peaked....

I would very much be interested to know the scientific/engineering basis for this statement.  Links?

Thanks.

--Jerome

[jsaliga]

As for Hard drive reliability - anyone who works in IT, and many who don't, are familiar with concepts such as redundant arrays of disks, mirroring and backup... - I use an array with dual redundancy - and I back it up to seperate offline disks periodically.

I personally don't use a RAID array for my music library.  I don't really require the fault tolerance that RAID provides, though I understand completely why someone would want to have their music library on a fault tolerant array of disks.

I require backups in the event of a disk failure.  Whenever I make a change to my live music library it is replicated to my online backup volume, and then the online backup is synchronized with an offline backup every week.  While my music library is very important to me, it can be down for a number of hours while I restore a failed disk from backup.  It isn't something that I require immediate access to 24/7.  Now my VMware vSphere cluster (yes, I have a three node VMware cluster at home) is a different beast.  For that I have a small SAN in a 0+1 RAID with a 10gb ethernet backbone.  That system is running about 40 servers in a lab and development environment that I need for work.  It has to be up and running 24/7 so I need fault tolerance for both disk and compute and also backups.

--Jerome

[Mag]

How do you know that to be the case? are you sure it isn't simple confirmation bias? - or is this a subjective opinion?

Well I did alter the sound to my liking with a mixer. But I have the proof for comparison if you are ever in the neighborhood, warning I'm a slob.

In one case for example I took an mp3 download and elevated it to a high quality recording. Another case where upsampling fixed or corrected distortions in the recording. So I know that it works and I have like 90 gigs worth of my favorites songs/albums as proof. A recording though has to have an element of transparency to work. It won't get rid of bad compression. With a good recording the differences are more subtle but noticable with careful listening.

[charmerci]

To address the initial question, in having all my music downloaded allows me to use shuffle and create my own "radio station" that can continuously play on.

[dlaloum]

Well I did alter the sound to my liking with a mixer. But I have the proof for comparison if you are ever in the neighborhood, warning I'm a slob.

In one case for example I took an mp3 download and elevated it to a high quality recording. Another case where upsampling fixed or corrected distortions in the recording. So I know that it works and I have like 90 gigs worth of my favorites songs/albums as proof. A recording though has to have an element of transparency to work. It won't get rid of bad compression. With a good recording the differences are more subtle but noticable with careful listening.

An MP3 is not a valid starting point! - It is a lossy storage format - that means when an MP3 is created from a standard (nonlossy) digital format - it is processed, and things that the algorhythm believe would not be noticed, are discarded. - These cannot be re-created.

Various upsampling methods and bits of software try to artificially generate harmonics and other material to provide a more pleasing "upsampled" version of the MP3 - these are further artificial additions.

Like the MP3 they differ from the original too ... you may like it or not, you may consider it an improvement, or not - but the end result is getting even further away from the original recording!

There are non-lossy compression methods and file formats (eg: flac, and others)

If you want an indication of the capabilities and potential quality / transparency capabilities of digital recordings, you have to start with a lossless format. Then the quality of the recording from LP, including everything from cartridge, loading, phono stage, turntable, through to ADC (digital converter) quality, etc... can be auditioned and/or measured.

I do have some high bandwidth mp3 recordings that are very good - but that is simply because at the time of purchase no alternative (ie: lossless format) was available - I don't use MP3 to archive/store my vinyl, as that would defeat its purpose! (IMO)

[jsaliga]

An MP3 is not a valid starting point!

It is for people who don't share your values and priorities. 

That being said, this is not something I would embrace for myself.  In the early 2000s I had a collection of about 500 albums in mp3 format.  At some point I decided to delete them all from my music library.  My goals and priorities shifted and I moved on to lossless formats and high resolution digital audio.

Why not just keep the mp3s I had instead of deleting them?  They don't take up a lot of space, after all.  Storage wasn't even a consideration.  I axed them because I was building a very large library of music and the likelihood that I would want to pull for one of those mp3 albums was so absurdly low that I decided that I wouldn't miss them at all.  I currently have a digital music library of about 5,500 albums, all lossless and mostly in high resolution audio, and a vinyl rack with roughly 1,100 LPs -- most of those are 180g and 200g audiophile pressings.  What on earth would I want with an mp3?  But the simple fact is that many folks are still more than satisfied with lossy compression.  And there are quite a few double blind tests out there that show the average listener cannot tell the difference between lossless and lossy compression.  It has been my experience that folks will turn a deaf ear to me when I suggest they can enjoy better quality sound by going lossless. 

--Jerome

[Mag]

An MP3 is not a valid starting point! - It is a lossy storage format - that means when an MP3 is created from a standard (nonlossy) digital format - it is processed, and things that the algorhythm believe would not be noticed, are discarded. - These cannot be re-created.

Various upsampling methods and bits of software try to artificially generate harmonics and other material to provide a more pleasing "upsampled" version of the MP3 - these are further artificial additions.

Like the MP3 they differ from the original too ... you may like it or not, you may consider it an improvement, or not - but the end result is getting even further away from the original recording!

There are non-lossy compression methods and file formats (eg: flac, and others)

If you want an indication of the capabilities and potential quality / transparency capabilities of digital recordings, you have to start with a lossless format. Then the quality of the recording from LP, including everything from cartridge, loading, phono stage, turntable, through to ADC (digital converter) quality, etc... can be auditioned and/or measured.

I do have some high bandwidth mp3 recordings that are very good - but that is simply because at the time of purchase no alternative (ie: lossless format) was available - I don't use MP3 to archive/store my vinyl, as that would defeat its purpose! (IMO)

 I've been using this method since 2005. I wanted to make a music cd for a prize at work for I think it was a Kaiser tournament or could have been for a Chili recipe, don't recall for sure. Anyway the music I wanted to use wasn't very good quality so I tried using a dac and recording it. I repeated this process like 5 times and it worked. Since about 2012 I've been able to speed up the process using an external dac and piggy backing on the dac in my pre-processor, but it's still time consuming.

I've also taken uncompressed blu-ray and it elevated that slightly as well.