• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Need your impression...

Which one is best for you? And why?

  • Total voters
    18
PCM because it sounds the same as DSD (a solution to a problem nobody has anymore) and is more flexible / compatible.

Regarding the linked post from 5 years ago: If there's an audible problem with the filters in your DAC, your DAC is garbage. If they're saying that most DACs have audible flaws in their filters, well... digital filters are very easy to ABX in Foobar, so let's see it?
 
Last edited:
kemmler3D said:
PCM because it sounds the same as DSD (a solution to a problem nobody has anymore) and is more flexible / compatible.

Regarding the linked post from 5 years ago: If there's an audible problem with the filters in your DAC, your DAC is garbage. If they're saying that most DACs have audible flaws in their filters, well... digital filters are very easy to ABX in Foobar, so let's see it?
Click to expand...


Participants were asked to choose which of the two stimuli has the higher sensation or impression in a given attribute. Therefore, DSD was chosen to have higher quality and preference in most of the attributes than PCM.
 
PaperBoat said:
Participants were asked to choose which of the two stimuli has the higher sensation or impression in a given attribute. Therefore, DSD was chosen to have higher quality and preference in most of the attributes than PCM.
Click to expand...
What jumped out at me reading the article was the PCM response rolled off some in lower frequencies vs DSD in the same Tascam machine. And sure enough, they found no discernment of sources that were high frequencies only, like the triangle. Seems pretty obvious, and the authors mentioned it in the paper that it is quite possible that low frequency difference is what they heard. This needs to be replicated with a unit that has the same response in PCM as DSD. Certainly plenty of ADC's are flat right down into lower frequencies. Why the Tascam is not is a bit of a good question. I wonder if any of the tests they didn't use from the other two interfaces showed anything at all?
 
Curvature said:
Flawed test.

Forced AB choice for certain attributes. The data is effectively useless.
Click to expand...
I don't think a forced choice makes it flawed. That is a common and well known method. More common in the food tasting industry. They use it just the way this test did. For instance, sample A and B pick the one which is more salty? Or more crunchy or which has the preferred mouth feel? You have to choose one, you aren't allowed to choose no difference. Read up on 2AFC testing. Two alternative forced choice.
 
Blumlein 88 said:
I don't think a forced choice makes it flawed. That is a common and well known method. More common in the food tasting industry. They use it just the way this test did. For instance, sample A and B pick the one which is more salty? Or more crunchy or which has the preferred mouth feel? You have to choose one, you aren't allowed to chose no difference.
Click to expand...
It doesn't make sense when you have PCM vs. DSD and are asking questions about spaciousness, timbre, etc. with limited signal measurements...
 
Curvature said:
It doesn't make sense when you have PCM vs. DSD and are asking questions about spaciousness, timbre, etc. with limited signal measurements...
Click to expand...
Why not? You are testing if those perceptions are different and trying to determine if the difference is reliably different between them.

Read these example of various test procedures. The description of the 2AFC is following precisely this method used in the paper.

In this paper I think the low end roll off of the PCM is what is being heard. You could test that if you could EQ the DSD to have that same response and repeat the test with EQ'd DSD vs straight DSD.
 
Blumlein 88 said:
Why not? You are testing if those perceptions are different and trying to determine if the difference is reliably different between them.

Read these example of various test procedures. The description of the 2AFC is following precisely this method used in the paper.

In this paper I think the low end roll off of the PCM is what is being heard. You could test that if you could EQ the DSD to have that same response and repeat the test with EQ'd DSD vs straight DSD.
Click to expand...

Blumlein 88 said:
Why not? You are testing if those perceptions are different and trying to determine if the difference is reliably different between them.

Read these example of various test procedures. The description of the 2AFC is following precisely this method used in the paper.

In this paper I think the low end roll off of the PCM is what is being heard. You could test that if you could EQ the DSD to have that same response and repeat the test with EQ'd DSD vs straight DSD.
Click to expand...
I'm familiar with forced choice procedures.

I disagree with your conclusions, but I don't have time right now to give a fuller response.
 
PaperBoat said:
Participants were asked to choose which of the two stimuli has the higher sensation or impression in a given attribute. Therefore, DSD was chosen to have higher quality and preference in most of the attributes than PCM.
Click to expand...
This is much more interesting than most evidence put forth for audible differences between formats. However, I'm a bit troubled by the lack of controls (they didn't compare the same format to itself to make sure the test setup was working properly) and although the recordings OUGHT to be equivalent due to the setup they described, they didn't actually confirm that anywhere in the paper. They level matched at 1khz but why not compare the spectra of the recordings anyway?

Given that they knew the recorders had different frequency responses (maybe 0.5dB in the audible band) that seems like something they should have corrected in the mixing desk. It calls into question whether there is a difference in the formats or just the recording equipment.

They also didn't seem to show the raw data, just asserted some p-values. So I guess we can't actually check if those calculations are correct. Oh well. Anyone know if the result has been replicated anywhere?

Anyway, that's that paper.

The comment linked previously talked about upsampling to DSD to avoid DSP, which has nothing to do with the experiment. It's also rare to see a recording made in this fashion (recorded straight to DSD with no digital processing) so it's not really applicable to home listening. If these results were replicated (big "if", maybe) then it might be worth considering at the industry / recording studio level. Until then it's neither here nor there for the home listener, who has low-to-no access to pure DSD recordings either way.
 
I'll say it again,sometimes maybe has to do with how the devices handle it,take a look for example how Khadas Tone1 does it with some of my rough measurements (grain of salt required,Amir's measurements for the same DAC is 110db SINAD at 1Khz so subtract accordingly):


www.audiosciencereview.com

PCM vs DSD

Did a little comparison and I was wondering if any friend with better gear here has tested this. Results you're about to see is probably narrowed to the specific DAC but it's interesting to see how it handles things differently. I will not try to decipher the results other that DSD seems much...
www.audiosciencereview.com www.audiosciencereview.com

I urge folks with proper gear to make similar measurements,I'm curious!
 
kemmler3D said:
It's also rare to see a recording made in this fashion (recorded straight to DSD with no digital processing) so it's not really applicable to home listening. If these results were replicated (big "if", maybe) then it might be worth considering at the industry / recording studio level. Until then it's neither here nor there for the home listener, who has low-to-no access to pure DSD recordings either way.
Click to expand...
As many of us know how recording houses are doing the dirty job... After capturing the signal in analogue or DSD they mixing it in PCM then master in DSD! I know that there are limited options for mixing in DSD native. But I'm curious to know what is the logic behind it? How a DSD content could be called as "DSD" after loosing its veracity?
 
Tried both PCM and DSD recordings on my system. There is no real difference to me. Sound is more dependent on what music and how it was recorded. So for everyday usage I stay mainly with PCM like it is on CD but if available 24-bit/96k. Perhaps with a very dedicated DSD only DAC it may sound a little better. Further I buy that DSD to analog conversion is relatively easy compared to PCM DAC, but to me a conversion analog from microphone or musical instrument to DSD may be not really easy and perfect.
 



This helps me a lot... ↓
 
Last edited:
PaperBoat said:



This helps me a lot... ↓
Click to expand...
Is there a reason you are so interested in upsampling to DSD? You raised 2 very similar threads on this in the space of a couple of days. Even if there is any discernible difference, do you really think it is going to be of the order that it actually makes a meaningful difference to your day to day listening?
 
Jimbob54 said:
Is there a reason you are so interested in upsampling to DSD? You raised 2 very similar threads on this in the space of a couple of days. Even if there is any discernible difference, do you really think it is going to be of the order that it actually makes a meaningful difference to your day to day listening?
Click to expand...
I'm just interested in learning... Nothing else. :)
 
You must log in or register to reply here.

Similar threads

MaxwellsEq
What am I missing - up-sampling and DSD conversion?
Replies
18
Views
2K
ZolaIII
ZolaIII
A
Strange DAP / DAC Foepoe UD951C - DSD Multichannel support?
2
Replies
38
Views
4K
Carznmusic
C
J
SACD with ESS chips. DSD or PCM?
Replies
6
Views
1K
Scytales
Scytales
Scytales
Is DSD HF Quantization Noise a Real Concern? A case study with modest measurements and a comparative survey.
Replies
16
Views
2K
dualazmak
dualazmak
NTTY
Denon DCD-SA1 Review (CD/SACD Player)
7 8 9
Replies
163
Views
26K
Sokel
Sokel
Back
Top Bottom