Hope this help you to explain Hi-Res music to your CD friends
- Thread starter sunjam
- Start date
- Status
Standard HP desktop computer with built-in headphone output. 16-bit, 44.1 kHz, 10,000 Hz, 0.5 V, sine wave. Analog oscilloscope. No step!
Last edited:
gregorio
Headphoneus Supremus
- Joined
- Feb 14, 2008
- Posts
- 6,878
- Likes
- 4,120
The basic premise of the article is: “To me, critical thinking and good learning technique can help us to be immune from these pseudo science virus.” - Which obviously here in the sound science discussion forum we would agree with. Unfortunately though, the linked article demonstrates the application of neither critical thinking nor a good learning technique and therefore accomplishes the exact opposite of “help us to be immune from pseudo science” and instead actually promotes pseudo science!!
It’s difficult to tell for sure if this article was deliberately intended to mislead/misinform others or if it’s genuine and the author is just seriously misinformed. As it appears to be the former, I won’t waste my time refuting it.
G
castleofargh
Sound Science Forum Moderator
- Joined
- Jul 2, 2011
- Posts
- 10,466
- Likes
- 6,102
The actual source for the measurements is linked as https://addictedtoaudio.com.au/blogs/how-to/how-to-pick-the-best-filter-setting-for-your-dac
Contrary to the link from OP, in the original article the guy understands what he's talking about and at no point does he make the silly silly claims or draw the silly silly conclusions of OP's link.
The scary oscilloscope screenshots seem to come from NOS DAC, and inappropriate filter or even no filter(depending on the DAC discussed). Also, the signal amplitude was picked to give the more dramatic(but perfectly expected) results, and that is mentioned in the original article.
Contrary to the link from OP, in the original article the guy understands what he's talking about and at no point does he make the silly silly claims or draw the silly silly conclusions of OP's link.
The scary oscilloscope screenshots seem to come from NOS DAC, and inappropriate filter or even no filter(depending on the DAC discussed). Also, the signal amplitude was picked to give the more dramatic(but perfectly expected) results, and that is mentioned in the original article.
sunjam
100+ Head-Fier
Thanks for your reminder of the link. The actual source of the pictures is indeed included in the blog. It shows clearly where the pictures come from. Of course, it would be great as you just reposted it here for the sake of convenience. Thanks again.
Could you share with us what do you mean by silly silly claims / conclusions? Did the blog give out wrong / incorrect info? If not, what's silly about it?
Mind to share with us how the article demonstrates the application of neither critical thinking nor a good learning? It would be a great example so that we all can learn from its mistakes (if there is indeed any).
Last edited:
gregorio
Headphoneus Supremus
- Joined
- Feb 14, 2008
- Posts
- 6,878
- Likes
- 4,120
OK, I’ll assume it was just a genuine misunderstanding due to the article itself being misleading:
A basic tenet of digital audio (as set out in the Nyquist-Shannon Sampling Theorem) is that the conversion from analogue to digital and from digital to analogue is band limited, IE. There are two filters, an anti-alias filter when converting to digital and an anti-image (or reconstruction) filter when converting from digital back to analogue. Uniquely, NOS DACs typically do not have a reconstruction filter and are therefore effectively broken! For this reason they are not used for professional or commercial audio applications, they are purely an audiophile marketing gimmick. Fortunately, NOS DACs are extremely rare, although in recent years the trend in the audiophile DAC market has been to provide switchable filter options and a few of them provide an option that emulates a NOS DAC. Though why anyone would want to emulate a broken DAC design in a DAC that isn’t broken is a mystery only an audiophile marketer (or someone suckered by it) could appreciate.
The article to which you linked does not demonstrate “critical thinking or good learning” because it doesn’t consider how DACs actually work, it only considers this rare, effectively broken design and therefore promotes the false marketing (pseudo science) which it falsely claims it’s trying to help others to avoid! By the author’s own words: “My critical thinking does not allow me to accept any claim without …further analysis.” - How is it possible that his “further analysis” only revealed a very rare, broken design but not how DACs are supposed to work or indeed how nearly all DACs do work? That is NOT critical thinking and it is DEFINITELY NOT a “good learning technique” if he hasn’t even learned the fundamental basics of how digital audio works! However as mentioned previously, he might actually have learned the basics and have applied critical thinking but is simply lying, in order to deliberately mislead others.
G
castleofargh
Sound Science Forum Moderator
- Joined
- Jul 2, 2011
- Posts
- 10,466
- Likes
- 6,102
Maybe I should have said blatant misrepresentation instead? When you have to go for non oversampling and inappropriate or no filter to demonstrate why hires is better, and show that signals have staircases, I feel like you're actually making a strong case for 44.1kHz and that old Monty's video.
Yes, NOS DACs with no filter exist, but then again, so do truck nuts and healing crystals. They are an insignificant portion of the DAC market and cannot be used to make a general case for anything. It's even less appropriate because applying a proper filter is an inherent part of signal reconstruction. It's like discussing how reinforced concrete behaves by showing a case of concrete without reinforcing bars.
It is objectively wrong to not filter the output of a DAC! A NOS DAC that's not complete nonsense, will have a filter. The most logical to me and fairly standard is a filter that rolls off early within the audible range(sometimes EQing the treble in a clearly noticeable way), as a way to get more attenuation in the out of band signal. No or bad filter means more aliasing and is objectively bad. Showing that is not the same as showing proper DAC behavior.
Why people do things is a question you'll have to ask them. DACs nowadays do more often offer a setting for a few filter options because those options are available in the chips to adapt to a particular design(there are often also some EQ solution, and it's really up to the designer to make those options available to the consumer). Once the design is done, there is IMO, no reason to change filters. But it's there and someone feels like there's a demand. So, why not? I don't care much for those myself.
As for that graph:
It comes from:
Electrical and Flow Characteristics of a Double-side Sliding Pulsed Discharge Plasma Actuator
talk about misrepresenting data.But yes, when you do absolutely everything to ruin digital reconstruction with all the bad choices available to you with non filtered NOS options, then playing hires improves the output fidelity. Not so much because it's hires, but because a higher sample rate behaves like an oversampled signal, we ask the file to do a little of what we denied for the DAC. Also with a bigger bandwidth, the unfiltered crap starts at higher frequencies and will hopefully not cause as much mess in the audible band.
The entire rational can be seen as finding the best way to fix a problem we deliberately created ourselves in the first place.
As for the conclusion,
There is in fact some truth to it. If I take my DACs, I think all of them measure a little better when I output a 24bit signal. But they also do if it's zero padding of a 16bit signal. So I just set my computer to output 24bit.
As for higher sample rate, things are more complicated IMO. There is often some benefit like how I mentioned, the filter has to effectively attenuate higher frequencies, so if it starts early, the attenuation will be stronger where the crap starts. But there could just as well be circumstances where the hires file brings a lot of ultrasonic crap that didn't exist on a redbook file. And we never know how the playback system will handle that.
Also, resampling/upsampling is used for reclocking/anti jitter or whatever clever tricks they have come up with in a chip. Sending higher sample rate could mean less to work with in those areas for the DAC.
Then nowadays there are all the DSD options.
I'm writing on the top of my head, and already I feel like the conclusion should be "it's complicated" and not some pro hires message that seems pulled out of a hat and backed up by out of context fringe examples.
For clarity that is your blog yes ?
sunjam
100+ Head-Fier
Cool, thanks a lot for your reply. I love to see people who demonstrate critical thinking.
Let's focus on a few points:
1. NOS DAC typically do not have a reconstruction filter?
No, they have analog filter. Did you have a chance to look at vintage NOS DAC? Do they ALL broken?
Even modern DAC have the analogy filter (in addition to the digital filter).
2. Many modern DACs support NOS mode (via NOS filter). All the flag ship Cirrus Logic chips and AKM chips support the NOS filter at the chip level.
3. Real NOS DAC (not using NOS emulation are getting more and more popular now). People are looking for these real NOS DAC to support the way how they play music.
4. Without the relevant knowledge, a lot of things would be considered as mystery. NOS DAC is a perfect example. Why people want NOS DAC in 2024? It is because they use the NOS DAC to playback 768k Hz or even higher bitrate (1536k Hz) music.
With my explanation, I hope it can give you a better view of the NOS DACs. Or you still consider NOS DACs are broken?
Here, specifically, we were talking about the NOS DAC without a filter...
Old NOS DACs used analog filters.
Filterless NOS DAC is the new BS...
Old NOS DACs used analog filters.
Filterless NOS DAC is the new BS...
bigshot
Headphoneus Supremus
See the article CD Sound Is All You Need in my sig. it’s all in there.
The benefits of high res are apparently in the way the brain processes the info after several minutes of listening, not in switching back and forth between formats and suddenly hearing something.
https://pubmed.ncbi.nlm.nih.gov/24722228/
bigshot
Headphoneus Supremus
It isn’t a benefit at all because even though it shows up in a brain scan, it does jack diddley for sound quality. There’s never been any evidence that super audible frequencies add anything at all to recorded music.
Yes, inaudible frequencies can have an effect on the body. They are used in sonic weapons to induce headaches and nausea. There’s more of an argument that they may contribute to listening fatigue.
Yes, inaudible frequencies can have an effect on the body. They are used in sonic weapons to induce headaches and nausea. There’s more of an argument that they may contribute to listening fatigue.
Last edited:
Ryokan
Headphoneus Supremus
I bought some hi-res music a few years back as 'everyone' was listening to hi-res files. After a while I got fed up with the space they took up, around 1.2gb each album, which adds up, so I deleted them and now only have the files in 320kbps. I couldn't hear a difference either. Which I'm glad about as they're often double the price, though sometimes on bandcamp you get a choice of mp3, flac and hi-res, plus a few others, all for the same price.
Ghoostknight
500+ Head-Fier
Which inaudible frequencies we speak about here?
i dont think you can have "portable" device that produces high volumes of subsonics.... so probably some ultrasonics?
i just read about sonic weapons that actually do this with volume, not frequencies, am i wrong?
Or do you speak about these teenager avoidance things, they do this with frequencies that teenager can hear, but usually not anyone above a certain age
Last edited:
- Status
Users who are viewing this thread
