The DAC Scam?

[captainsubtext]

Have heard the 2.x signature in my system. It didn't completely float my boat. It was better in same ways, worse in others.

AN 3.x might have been better but was completely beyond budget. I was looking to spend around 5 lacs and the T+A I ended up getting just was on a completely different league compared to the 2.x signature.

T+A with Hqplayer upsampling to dsd perhaps. I don’t find the PCM to be as good comparatively.

 

[Daniel.Moore]

That's a great point! The fact that we perceive differences between DACs, such as AKM vs ESS, suggests they aren't all the same. DAC measurements help us understand technical aspects like distortion, noise, and frequency response, but they don't always capture subjective listening experiences. While modern DACs may share some similarities in performance, nuances can arise from factors like implementation (e.g., power supply, output stage), which affect sound quality.


So, while measurements give us a technical foundation, personal taste and the context of how a DAC is integrated into a system can lead to differences in perceived sound. Testing different DACs allows users to make informed decisions based on both measurements and their subjective preferences.

 

[reignofchaos]

T+A with Hqplayer upsampling to dsd perhaps. I don’t find the PCM to be as good comparatively.

PCM section is more forgiving than the DSD section - it's good for bad recordings. Even in PCM, you need to use NOS mode and upsample to 768kHz for it to sound its best. Also internal volume level needs to be reduced to around -16dB else it introduces distortion in loud sections.

However with hqp dsd upsampling, it is a completely different beast. It sounds sublime with good recordings! 95% of my consumption is in DSD mode with hqp.

 

[Kannan]

That's a great point! The fact that we perceive differences between DACs, such as AKM vs ESS, suggests they aren't all the same. DAC measurements help us understand technical aspects like distortion, noise, and frequency response, but they don't always capture subjective listening experiences. While modern DACs may share some similarities in performance, nuances can arise from factors like implementation (e.g., power supply, output stage), which affect sound quality.


So, while measurements give us a technical foundation, personal taste and the context of how a DAC is integrated into a system can lead to differences in perceived sound. Testing different DACs allows users to make informed decisions based on both measurements and their subjective preferences.

Yesterday while deciding on DACs for my friend's newly acquired speakers, we did a back to back session with a DAC with AKM velvet chip and another with ESS9038pro. The filters and gains were perfectly matched and a real time back-to-back listening using an amplifier with multiple inputs showed absolutely no difference in sound signature.

 

[navbali]

I think a better DAC also needs a better chain to shine, having a high-end DAC with other electronics and speakers not up to mark will not reveal anything

 

[reignofchaos]

Yesterday while deciding on DACs for my friend's newly acquired speakers, we did a back to back session with a DAC with AKM velvet chip and another with ESS9038pro. The filters and gains were perfectly matched and a real time back-to-back listening using an amplifier with multiple inputs showed absolutely no difference in sound signature.

Which DACs? A modern oversampling DS chip barely has any sound - it is always the ancillary circuitry around that imparts a flavor.

 

[square_wave]

Yesterday while deciding on DACs for my friend's newly acquired speakers, we did a back to back session with a DAC with AKM velvet chip and another with ESS9038pro. The filters and gains were perfectly matched and a real time back-to-back listening using an amplifier with multiple inputs showed absolutely no difference in sound signature.

Can you tell us a bit more about the testing process?

What did you listen to and What were you looking for? Full songs or small samples ? Simple music or complex ?

What is the overall system ? And how is it setup ?

 

[captainsubtext]

Yesterday while deciding on DACs for my friend's newly acquired speakers, we did a back to back session with a DAC with AKM velvet chip and another with ESS9038pro. The filters and gains were perfectly matched and a real time back-to-back listening using an amplifier with multiple inputs showed absolutely no difference in sound signature.

What was the chain? DACs with the same chip can sound very different based on the implementation/power supplies etc. Headphones are an easier way to test because you're taking the room out of th equation

 

[drkrack]

absolutely no difference in sound signature.

That's a Good kind of System to have, ask him not to go about listening around for other DACs. Unheard Details getting revealed may not be a good option for sometimes.

 

[square_wave]

That's a Good kind of System to have, ask him not to go about listening around for other DACs. Unheard Details getting revealed may not be a good option for sometimes.

I am quite surprised to hear an AKM based DAC sounding the same as an ESS. Someone I know who made a product out of ESS Sabre spend almost a year on R&D and design iterations after the "almost final prototype" just to get the ESS to sound like analogue. He was trying to leverage the detail and resolution retrieval capabilities of ESS but at the same time remain musical. His previous model used a burr brown chip and task was so much easier.

 

[Kannan]

The speaker was Monitor Audio Platinum. The amplifier was Audiolab 8000A in power amp mode. The DACs were Topping D50 iii (ESS DAC) and Topping E70 Velvet (AKM). The preamps of both DACs was employed. Each output was connected to one of the inputs of the amp so that we could switch in real time. Both outputs were perfectly gain matched and employed the same filter.
Both DACs were connected to laptops via USB.

 

[square_wave]

PCM section is more forgiving than the DSD section - it's good for bad recordings. Even in PCM, you need to use NOS mode and upsample to 768kHz for it to sound its best. Also internal volume level needs to be reduced to around -16dB else it introduces distortion in loud sections.

However with hqp dsd upsampling, it is a completely different beast. It sounds sublime with good recordings! 95% of my consumption is in DSD mode with hqp.

I am new to HQ player although I have heard about its upsampling capabilties.

If you convert hqp dsd upsamplng, I presume you need a dac that can decode dsd, no? Or is there some fallback option ?

" Even in PCM, you need to use NOS mode and upsample to 768kHz for it to sound its best. "

What does this mean ? NOS mode means no over sampling , no ?

Also, this HQ player thing -- Is it better than all the usual streamer solutions ? I mean the high end ones?

 

[reignofchaos]

I am new to HQ player although I have heard about its upsampling capabilties.

If you convert hqp dsd upsamplng, I presume you need a dac that can decode dsd, no? Or is there some fallback option ?



What does this mean ? NOS mode means no over sampling , no ?

Also, this HQ player thing -- Is it better than all the usual streamer solutions ? I mean the high end ones?

You can upsample either to PCM or DSD depending on the capabilities of your dac. For DSD, 1 bit DACs based on vector step ladders are recommended such as T+A, Holo, Audiobyte etc.

The primary requirement is that the DAC have a true NOS (non oversampling) mode where it is possible to disable all digital filtering and digital processing on the device. If this doesn't exist, hqplayer is of no use.

 

[reignofchaos]

The speaker was Monitor Audio Platinum. The amplifier was Audiolab 8000A in power amp mode. The DACs were Topping D50 iii (ESS DAC) and Topping E70 Velvet (AKM). The preamps of both DACs was employed. Each output was connected to one of the inputs of the amp so that we could switch in real time. Both outputs were perfectly gain matched and employed the same filter.
Both DACs were connected to laptops via USB.

These topping DACs probably have the exact same power supply, input filtering and output stage. It will be nigh impossible to distinguish between them.

 

[Kannan]

These topping DACs probably have the exact same power supply, input filtering and output stage. It will be nigh impossible to distinguish between them.

Good enough, shows to go, the company knows what is is doing.

 

[square_wave]

You can upsample either to PCM or DSD depending on the capabilities of your dac. For DSD, 1 bit DACs based on vector step ladders are recommended such as T+A, Holo, Audiobyte etc.

Alright!

The primary requirement is that the DAC have a true NOS (non oversampling) mode where it is possible to disable all digital filtering and digital processing on the device. If this doesn't exist, hqplayer is of no use.

So, this means all the over sampling and processing happens in the player and the dac itself does no damage ? Why is oversampling in a player better than what you find in a dac itself ? Just curious. Maybe I am missing something.

 

[reignofchaos]

Alright!



So, this means all the over sampling and processing happens in the player and the dac itself does no damage ? Why is oversampling in a player better than what you find in a dac itself ? Just curious. Maybe I am missing something.

Yep! Exactly that is what is happening. Upsampling/Oversampling on a PC is far superior as one can use way more advanced algorithms due to available compute compared to the basic algorithms baked into the chips or an FPGA. Downside is the need to have a powerful PC.

 

[captainsubtext]

Yep! Exactly that is what is happening. Upsampling/Oversampling on a PC is far superior as one can use way more advanced algorithms due to available compute compared to the basic algorithms baked into the chips or an FPGA. Downside is the need to have a powerful PC.

For PCM upsampling you don’t need a very powerful PC.

For DSD, the modulator matters more than the sample rate and for the nicer modulators you do need some serious grunt.

 

[drkrack]

Yep! Exactly that is what is happening. Upsampling/Oversampling on a PC is far superior as one can use way more advanced algorithms due to available compute compared to the basic algorithms baked into the chips or an FPGA. Downside is the need to have a powerful PC.

Agree
I have 3 locations where I can up sample, at Lumin, Roon Server and Denafrips DAC.
Up Sampling at Roon Server works best, it's i5 11th Gen with 16gb Ram, it was worst when I Ran Roon on Synology NAS 423+, Better with i5 8th Gen Nuc , but current i5 11th Gen Nuc makes A pretty good difference. The dynamics used to be lost for simple dsp like volume leveling with Celeron of NAS.

 

[reignofchaos]

For PCM upsampling you don’t need a very powerful PC.

For DSD, the modulator matters more than the sample rate and for the nicer modulators you do need some serious grunt.

Yes agree - PCM upsampling can be done even by a 10 year old CPU. PCM upsampling in hqplayer is definitely better than upsampling within the DAC.