To upsample or not upsample, that is the question

Audiophile bill

New member
Joined
Jul 3, 2015
Messages
144
Hi all,

Based on the other thread for the golden gate and my own experience with my balanced big 7, and upsampling with various software, I think it would be great to start a latest state of play on the issue.

My personal conclusion using hqplayer and foobar to upsample all PCM to dsd 256 is that this route sounds much better to native PCM via hqplayer, foobar, jriver, or cd via oppo spdif.

What aspects sound superior to me?

> Resolution is improved - I hear more ambient cues and finest details
> Tangibility of the performer - sounds like they are even more there with me in the room
> Timbre of instruments is more like their native sound to my ears - specifically strings where the upsampling removes that tiniest digital edge that precludes the harmonics coming through
> Soundstage is clearer


It must also be noted that different upsampling algorithms sound quite different but overall it is much better.

So - let's discuss experiences again on this subject.
 
So are you saying even when ripping CDs, lets say with JRiver, up sampling them to DSD sounds better than doing a straight FLAC rip? I was under the assumption that the rip could not be better than the original source, the CD in this example.

Which leads me to another question, at this point I do not have true DSD playback. My McIntosh only has PCM with a maximum of 192 kHz. Therefore to playback a DSD file JRiver then down samples to 192 kHz. Therefore would it be better to rip to DSD, up sampled, and then playback and have JRiver down sample back to my McIntosh DAC?

Or does your above statement only hold true if played back through a true DSD decoder?

I might have to try a sample of this. Rip a real nice recorded CD, lets say Pink Floyd The Endless River into DSD and play it back down sampled. Comparing the original CD to the FLAC rip I can not tell any difference. While comparing the CD to a HD recording I hear a huge difference.
 
Going from PCM to DSD causes infinitely less problems. If you have a best in class DSD playback engine and upsample PCM to DSD256 on an optimised Server running in Core Mode and plyed back thru Bughead Emperor shareware?? Dynamite!
You would need something like a Trinty Dac at $50K to get native PCM to equal that. Perhaps the Phasure NOS1a could do that about as well too, but that is a PITA to operate…but reportedly very dynamic.
 
hmmm.... I just experimented. Not at all scientific. And I tested through my headphones so not really testing sound stage and imagine, etc. I ripped the Pink Floyd CD to DSD256. I played the CD back and the DSD through JRiver being down sampled to 192 kHz.

Comparing back and forth and I can't believe I am writing this, but I swear I hear a difference. The DSD rip seemed a little bit smoother, more open, better dynamic range... like it had more of the warmth of analog. While the CD sounded more.... digitally... if that even makes any sense. Very subtle difference, but I swear it was there.

How is that possible. When JRiver up scales it is it able to add some of that analog warmth back in :huh:... I would not think that was possible especially since JRiver is down scaling it on playback...

Certainly worth trying a few more rips and of course listening through the speakers....
 
Ok, I have been doing some serious back and forth listening and finally believe, that any difference I was hearing was an anomaly and that there is no improvements ripping to DSD and then down converting the file back to 192 kHz for playback.

Also when doing a ton of research on this subject it has become apparent that all DSD files have traveled through a PCM stage at one point or another. No mastering edits can be done on a DSD file.

The more and more I read the more it has become evident that DSD "could" be better, but in it's current format it is not. If everything was recorded from the very get go in DSD, and keep in DSD throughout the chain than it would be better. However every commercially available recording has gone through transition to and from PCM throughout it's life. Every step adds distortion, period.

So this is my discovery process, take it as you will.
 
I have been doing some serious back and forth listening and finally believe, that any difference I was hearing was an anomaly and that there is no improvements ripping to DSD and then down converting the file back to 192 kHz for playback.

Not understanding why you would ever consider that this process may result in an "improvement."
 
I don't know.... an experiment from something I read.... at first I thought it might be an improvement, but with more test, not so much....
 
Ok, I have been doing some serious back and forth listening and finally believe, that any difference I was hearing was an anomaly and that there is no improvements ripping to DSD and then down converting the file back to 192 kHz for playback.

Also when doing a ton of research on this subject it has become apparent that all DSD files have traveled through a PCM stage at one point or another. No mastering edits can be done on a DSD file.

The more and more I read the more it has become evident that DSD "could" be better, but in it's current format it is not. If everything was recorded from the very get go in DSD, and keep in DSD throughout the chain than it would be better. However every commercially available recording has gone through transition to and from PCM throughout it's life. Every step adds distortion, period.

So this is my discovery process, take it as you will.
This has been discussed to death and I dont have the appetite to point out where you are wrong. You read one side of the debae and assume you have all the answers…well then fine.

A man convinced against his will, is of the same opinion still.
 
I have no side what so ever. Just reading every article I can. No where what so ever did I say I have all the answers, in fact I have stated several times the exact opposite. I have tried to find and have found many articles, on both sides as you say. Unfortunately a vast majority state as I wrote.

It is like you take it as a personal insult and then you proceed to be to make derogatory and insulting statements. Mellow out dude, it was simply a discussion; no one was insulting you... I wish the opposite could be said.
 
Randy - check native dsd downloads. Recorded natively in up to quad dsd - it is special.

Everything I have read Bill, this definitely sounds like it is the case. At some point I want to do precisely this!

One question is how would I tell if something was a native DSD recording, versus being up-sampled. I assume most music from the 1970/80, pop, classic rock, folk, folk rock, etc., was not originally recorded in DSD level. Are there popular type recordings in DSD? Pink Floyd, Fleetwood Mac, Shawn Phillips, etc. I believe I have seen someone claiming that many of these were recorded in DSD level. I found this hard to believe. However I have purchased many re-mastered versions in 96 kHz and 192 kHz which are fabulous.
 
Everything I have read Bill, this definitely sounds like it is the case. At some point I want to do precisely this!

One question is how would I tell if something was a native DSD recording, versus being up-sampled. I assume most music from the 1970/80, pop, classic rock, folk, folk rock, etc., was not originally recorded in DSD level. Are there popular type recordings in DSD? Pink Floyd, Fleetwood Mac, Shawn Phillips, etc. I believe I have seen someone claiming that many of these were recorded in DSD level. I found this hard to believe. However I have purchased many re-mastered versions in 96 kHz and 192 kHz which are fabulous.

Hi Randy,

The rationale for my recommendation to check out native dsd website is that all their recordings are sourced either from dsd masters or analog tape. None of them come from PCM.

Many of the recordings of the era you describe would have been analog tape masters...
 
Well good... so this is something for me to look into for a future upgrade! Getting the Rel sub in (hopefully tomorrow) and then got to let things settle a little. Will probably consider adding a solid DSD unit for next upgrade. The new TEAC sounds pretty interesting.
 
I have no side what so ever. Just reading every article I can. No where what so ever did I say I have all the answers, in fact I have stated several times the exact opposite. I have tried to find and have found many articles, on both sides as you say. Unfortunately a vast majority state as I wrote.

It is like you take it as a personal insult and then you proceed to be to make derogatory and insulting statements. Mellow out dude, it was simply a discussion; no one was insulting you... I wish the opposite could be said.
You make definitive statements and act as if you KNOW, when clearly you don't.

You need to take a chill pill and ask and read more widely, as there is a lot of propaganda out there.

In addition to all you have digested negatively by some propagandists, is the fact that no matter HOW some albums were recorded, the ONLY hi-res version that is/was available was DSD on SACD disks. To rip that to DSD and then to further decimate back to hi-res PCM is not a good plan, given the logic you yourself put forward..every decimation causes loss. Indeed many postulate that the source of some high-fez PCM commercially is either up sampled rebook of DSD decimation from PRO rips of the SACDs.
 
Just smile and wave.... smile and wave....

...as I said, I am on a learning curve and having discussions of the over whelming amount of articles I have read... and very much will read any article on either side... but please do not put words in my mouth.... thank you very much....
 
Ok, this is a complete moot issue. I just ordered a DSD DAC. Not a super high end one, but hopefully a decent one. I ordered a TEAC which from what I understand is associated with the Esoteric brand. At least it will give me a start. I have four DSD albums to use as my test subjects(Blood Sweat and Tears, Cat Stevens, and two Alan Parsons Project albums), :)... and if what you say is true I will be mainly getting my new music from the DSD choices. We will see....
 
So back to the original discussion. I noticed reading through the Teac manual that the unit can automatically up sample a PCM signal, as can JRiver.

Therefore the question, is it a benefit to up sample PCM or is it better to simply play them in their original format?
 
So back to the original discussion. I noticed reading through the Teac manual that the unit can automatically up sample a PCM signal, as can JRiver.

Therefore the question, is it a benefit to up sample PCM or is it better to simply play them in their original format?

Aha - well that was the title of my thread :)

There is not a lot of consensus right now as there are camps who believe that upsampling PCM to dsd is the way forward as long as it is done via hqplayer software (jriver is not good for upsampling, Foobar is less good by okay) and another camp that favour native playback.

I don't believe there is a definitive answer to this but based on my assessment - if you like the bite that PCM tends to bring or your system is a bit flat, then perhaps stick native. If you have enough life but enjoy the upper treble dsd sheen (as I call it) then this might be your way. I would try both and decide for yourself - I think it will be system dependent and user preference.
 
Yea, probably trial and error is the way. Another thing I noticed reading through the manual is that the Teac does on the fly up sampling.... May have to try that back and forth as well. It is a setting in the menu, and they do cover it within the manual.

Also, any good sites for getting DSD albums. I have gone through Acoustic Sounds selection. A few things there that I might want, but not a lot. If I was Bob Dylan fan... holy smokes I would be in hog heaven. I have my one Dylan album already and I'm good :)....
 
Back
Top