Upscaling music files resolution - worth it or a dream?
post-15694082
Thread Starter
Post #1 of 9

FenceFurniture

Head-Fier
Joined
Oct 4, 2018
Messages
91
Reaction score
44
Location
Sydney, Oz
Joined
Oct 4, 2018
Location
Sydney, Oz
Posts
91
Likes
44
Does anyone think that there is anything at all to be gained by upscaling from 16bit 44.1kHz to 24bit 176.4kHz?

I've just been listening to a very well recorded track which was originally 16-44 and it had been upscaled to 24-176. Swapping between both versions playing the same part of the same track (through a Schiit Jotunheim and Focal Stellias) and I couldn't discern any difference whatsoever. The upscaled track is 4x the size though!

(Mods - if this isn't in the correct area please move it).
 
     Share This Post       
post-15694106
Post #2 of 9

Roseval

500+ Head-Fier
Joined
Apr 12, 2009
Messages
964
Reaction score
210
Joined
Apr 12, 2009
Posts
964
Likes
210
Going from 16 to 24 bits won’t have any impact as adding 8 zero bits won’t affect sound quality.
Beside, as most DAC’s are 24 bit today, it will happen anyway.

Up-sampling won’t add any information either.
A 44 kHz can’t reproduce anything above 22 kHz.

Except if you have a NOS DAC, all other DACs do either up- or oversampling.
Doing up-sampling e.g. using a media player, means your audio is resampled twice.
Once by the media player and once by the DAC.
In principle you won’t have this.
In practice you won’t hear the difference.

IMHO best practice is to leave the original sample rate as it is.
 
     Share This Post       
post-15694148
Post #3 of 9

PurpleAngel

Headphoneus Supremus
Joined
Sep 1, 2011
Messages
28,702
Reaction score
1,752
Location
California
Joined
Sep 1, 2011
Location
California
Posts
28,702
Likes
1,752
Does anyone think that there is anything at all to be gained by upscaling from 16bit 44.1kHz to 24bit 176.4kHz?

I've just been listening to a very well recorded track which was originally 16-44 and it had been upscaled to 24-176. Swapping between both versions playing the same part of the same track (through a Schiit Jotunheim and Focal Stellias) and I couldn't discern any difference whatsoever. The upscaled track is 4x the size though!

(Mods - if this isn't in the correct area please move it).
No increase in audio quality.
 
     Share This Post       
post-15695527
Post #4 of 9

FenceFurniture

Head-Fier
Joined
Oct 4, 2018
Messages
91
Reaction score
44
Location
Sydney, Oz
Joined
Oct 4, 2018
Location
Sydney, Oz
Posts
91
Likes
44
Going from 16 to 24 bits won’t have any impact as adding 8 zero bits won’t affect sound quality.
That's the way I see it (but I'm not technically inclined).

It came up recently with someone I know who swears they can hear a difference - says it's a slightly smoother sound. Even if it is, I can't see that it's worth increasing the file size by 300%.

So I quoted the answers in this thread and the response was that unless you have an ESS Sabre DAC you won't hear the difference. (my Toshiba laptop - even though it's a 2016 model - has a DAC capable of only 48kHz, which seems quite remarkable these days). Reckons it's something to do with a particular type of distortion (which couldn't be brought to mind in our discussion).

Anyhoo, thanks for the responses - as I thought.
 
     Share This Post       
post-15696268
Post #6 of 9

bfreedma

The Hornet!
Joined
Feb 3, 2012
Messages
2,442
Reaction score
1,328
Joined
Feb 3, 2012
Posts
2,442
Likes
1,328
Depends on your point of views. Chord and M-Scaler would wage in differently
Of course they would - gotta sell that product...

I've looked on Chord's site many times for a viable technical explanation or testing showing audibility of their product's "improved" technology implementation - no luck finding anything yet.
 
     Share This Post       
post-15708071
Post #7 of 9

InstantSilence

500+ Head-Fier
Joined
May 1, 2020
Messages
506
Reaction score
141
Location
usa
Joined
May 1, 2020
Location
usa
Posts
506
Likes
141
Of course they would - gotta sell that product...

I've looked on Chord's site many times for a viable technical explanation or testing showing audibility of their product's "improved" technology implementation - no luck finding anything yet.
So with a resolving set up. You can definitely hear it, I tried with hqplayer and it truly makes a significant difference.
But with headphones and dacs under 2k you don't hear much as the resolution is budget
 
     Share This Post       
post-15708090
Post #8 of 9

FenceFurniture

Head-Fier
Joined
Oct 4, 2018
Messages
91
Reaction score
44
Location
Sydney, Oz
Joined
Oct 4, 2018
Location
Sydney, Oz
Posts
91
Likes
44
When I did the comparison it was with a pair of Focal Stellias (standard supplied balanced cable) coming out of Schiit Jotunheim DAC/Amp. Ok the Schiit isn't TOTL but the headphones are right up there.
 
     Share This Post       
post-15708516
Post #9 of 9

castleofargh

Sound Science Forum Moderator
Joined
Jul 2, 2011
Messages
9,079
Reaction score
4,416
Joined
Jul 2, 2011
Posts
9,079
Likes
4,416
That's the way I see it (but I'm not technically inclined).

It came up recently with someone I know who swears they can hear a difference - says it's a slightly smoother sound. Even if it is, I can't see that it's worth increasing the file size by 300%.

So I quoted the answers in this thread and the response was that unless you have an ESS Sabre DAC you won't hear the difference. (my Toshiba laptop - even though it's a 2016 model - has a DAC capable of only 48kHz, which seems quite remarkable these days). Reckons it's something to do with a particular type of distortion (which couldn't be brought to mind in our discussion).

Anyhoo, thanks for the responses - as I thought.
The thing is, modern DACs will oversample to maybe 10 times the rate of your file, often the final sample rate will be fixed and the calculation will be to turn whatever you feed into that specific sample rate the DAC is optimized for(filter and what not). So you can guess why it could be relevant and why it probably isn't. It has to do with your oversampling method pitted against the DAC's, or how much of that resampling was used as an anti jitter upsampling solution, and stuff like that. You'll meet cases where oversampling will give you improved final resolution out of the DAC, and cases where the opposite will be true and you would have done better to leave the DAC handle things the way it was supposed to. As to audibility. Well, I would assume it's rare, but I wouldn't claim it's impossible as I do not know the possible consequences for a given DAC. Most likely not something worth losing sleep about.

Another case worth considering comes from bad settings: I would bet many Pringles boxes that some people force a resampling but have Windows set to some other value so it resamples after they just did, and then that goes to the DAC that will most likely convert the all thing into a handful of bits at crazy high rate. So now we're not talking about oversampling but about several up and down sampling including one from the notoriously "meh" Windows converter(no idea about other platforms).

Last case is the person who owns some antiquated NOS R2R DAC: because those typically cause either a lot of aliasing, or role off the trebles(sometimes audibly so) when playing lower sample rates; it becomes very possible that oversampling the signal will improve the sound. Or at least change it in a noticeable way. That is something I've experienced personally once. So I know it can happen, but I don't know the likelihood of audible changes with such tech(rightfully a niche tech nowadays). There is some deep irony in a user purchasing a non oversampling DAC and then really fixing it by oversampling himself. but that's just my evil minded opinion😈.
 
     Share This Post       

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top