Multi-bit DSD versus PCM

[Blizzard]

At the end of the day, one can also upsample these to quad dsd in software and add filters as you see fit using hqplayer or other software to further "improve" the sound quality.

Yes but I've found the extra layer of filtering to DSD degrades the quality. You must process it through HQplayer's multibit SDM engine like is the topic of this thread in order to do this. It's much better to do it all in 1 shot and use the directSDM bypass in HQplayer for DSD. Unless of course you want to use this feature for room correction, crossovers for active speakers etc.

So if this procedure was done with an SDM chip like the Sabre, and all others without DSDdirect mode, you would be processing the DSD through 2 different multibit SDM engines before it get's to the analog outputs. Plus the initial 1 bit SRC/SDM. So 3 stages in total of SRC/SDM in the signal path, one single bit, plus 2 multibit.

 

[Audiophile Bill]

Yes but I've found the extra layer of filtering to DSD degrades the quality. You must process it through HQplayer's multibit SDM engine like is the topic of this thread in order to do this. It's much better to do it all in 1 shot and use the directSDM bypass in HQplayer for DSD. Unless of course you want to use this feature for room correction, crossovers for active speakers etc.

So if this procedure was done with an SDM chip like the Sabre, and all others without DSDdirect mode, you would be processing the DSD through 2 different multibit SDM engines before it get's to the analog outputs. Plus the initial 1 bit SRC/SDM. So 3 stages in total of SRC/SDM in the signal path, one single bit, plus 2 multibit.

Thanks - makes sense. The question I have though is whether these differences are noticeable and if so by what % roughly? It seems to me that if we are interested in making analogue recordings, then we should apportion most of our resource in the analogue rather than the ADC. Perhaps I am wrong but I honestly can't believe that changing my Tascam to a Horus using double dsd recording for both and using the Horus in the purest signal path sense as described that the actual file produced at the end will be much different and not easily distinguished under blind conditions. Spending the same dosh on a $5k cartridge upgrade and using the Tascam would surely produce the better file. So what I am saying is that we all operate with fixed resources other than the lucky few and thus I can't imagine spending much on the ADC side since I believe the money to be spent better on the source itself. Thoughts?

 

[Bruce B]

So really what Ian needs if he wants the holy grail, is a Merging HAPI ADC, combined with the Pyramix mastering software. This way he can make an outstanding vinyl rip collection that would blow away half of the commercial digital releases available. If I were him this would be a no brainer

 

[Bruce B]

Thanks - makes sense. The question I have though is whether these differences are noticeable and if so by what % roughly? It seems to me that if we are interested in making analogue recordings, then we should apportion most of our resource in the analogue rather than the ADC. Perhaps I am wrong but I honestly can't believe that changing my Tascam to a Horus using double dsd recording for both and using the Horus in the purest signal path sense as described that the actual file produced at the end will be much different and not easily distinguished under blind conditions. Spending the same dosh on a $5k cartridge upgrade and using the Tascam would surely produce the better file. So what I am saying is that we all operate with fixed resources other than the lucky few and thus I can't imagine spending much on the ADC side since I believe the money to be spent better on the source itself. Thoughts?

This goes back to what I've done for years and especially the Wilson Audiophile Recordings. I recorded the master tape transfers on 5-6 different ADC's. I then blindly put them up on the server and let the client choose which one they liked the best. There are pretty big differences between Korg/Tascam recorders and a Grimm/Horus.

 

[Blizzard]

Thanks - makes sense. The question I have though is whether these differences are noticeable and if so by what % roughly? It seems to me that if we are interested in making analogue recordings, then we should apportion most of our resource in the analogue rather than the ADC. Perhaps I am wrong but I honestly can't believe that changing my Tascam to a Horus using double dsd recording for both and using the Horus in the purest signal path sense as described that the actual file produced at the end will be much different and not easily distinguished under blind conditions. Spending the same dosh on a $5k cartridge upgrade and using the Tascam would surely produce the better file. So what I am saying is that we all operate with fixed resources other than the lucky few and thus I can't imagine spending much on the ADC side since I believe the money to be spent better on the source itself. Thoughts?

Yes it makes an audible difference. Enough for me that it's the only way I will ever listen to DSD now with passive speakers. Unless going active where the DSP on the DSD is a must. Then you have to decide what the lesser of the 2 compromises are, passive crossovers, or multibit SDM DSP. In this situation, the lesser of the 2 comprimises will likely be the multibit SDM. After all it's still the finest way possible to do DSP. Far beyond what's possible on the best Sharc DSP chips.

As for the money spent on the analog source, I find it far more cost effective to buy master tape DSD rips, guys like Bruce did from NativeDSD and Analogue sounds, and get my vinyl rips for free from my buddy with the $100K plus rig.

Try for yourself with HQplayer. When you play a native DSD track and DirectSDM isn't checked, it gets processed through the multibit SDM engine. When checked, it bypasses the multibit SDM engine. In most SDM DAC chips, you can't bypass this engine, but you can on some, but only with DSD. Your GG doesn't have the SDM/SRC processing in it. the DSD section is just a simple low pass filter. So if you send it DSD trough HQplayer with SDMdirect enabled, it doesn't go through any multibit processing. This is the purest path possible. In this scenario the only SRC/SDM the audio sees is the initial 1 bit from the original ADC/DAW. So what this shows is there's no substitute for using the best 1 bit DSD SDM/SRC algorithms the first time around. This can only be done with the original ADC/DAW package. This is another reason DSD is better that is overlooked by the PCM advocates. You can never have a signal path this pure from start to finish with PCM, unless you use an inferior R2R DAC. But that technology is outdated now. The last R2R chip the BB 1704 was unveiled back in 1998. The days of SDM tech, not living up to R2R standards are in the past now. The advanced SDM algorithms in DAW like Pyramix and HQplayer will put any R2R to shame.

 

[Blizzard]

This goes back to what I've done for years and especially the Wilson Audiophile Recordings. I recorded the master tape transfers on 5-6 different ADC's. I then blindly put them up on the server and let the client choose which one they liked the best. There are pretty big differences between Korg/Tascam recorders and a Grimm/Horus.


View attachment 25572

View attachment 25573

What would you know Bruce? You don't have the hands on experience

 

[Bruce B]

What would you know Bruce? You don't have the hands on experience

Yeah, I know I'm a hack at this.

 

[esldude]

You guys managed to completely ignore Audiophile Bill's question. If you had limited resources and had $5k to improve a system. Which is better, $5k better TT setup with Tascam doing the ADC duties or putting that toward a Merging Horus ADC (which actually would be more than $5k)?

A bit of an open ended question so maybe you need to frame it in a particular TT/system context.

We know $5k can improve an analog rig a good amount, so is the upgrade in DSD ADC going to be an even larger amount of improvement for the same money?

 

[Audiophile Bill]

You guys managed to completely ignore Audiophile Bill's question. If you had limited resources and had $5k to improve a system. Which is better, $5k better TT setup with Tascam doing the ADC duties or putting that toward a Merging Horus ADC (which actually would be more than $5k)?

A bit of an open ended question so maybe you need to frame it in a particular TT/system context.

We know $5k can improve an analog rig a good amount, so is the upgrade in DSD ADC going to be an even larger amount of improvement for the same money?

Thanks for reiterating the question

 

[Audiophile Bill]

So to reframe in my system context.

Turntable is Kuzma Ref with Triplanar VII and Lyra Kleos through BMC MCCI.

Option 1. Buy a Merging Horus and associated software for not sure price but let's just say $7k

Or

Optio. 2. Buy an Air Tight PC1 or Lyra Etna or whatever very high end cartridge. Use existing Tascam.

What recording sounds better? I reckon option 2.

 

[Blizzard]

You guys managed to completely ignore Audiophile Bill's question. If you had limited resources and had $5k to improve a system. Which is better, $5k better TT setup with Tascam doing the ADC duties or putting that toward a Merging Horus ADC (which actually would be more than $5k)?

A bit of an open ended question so maybe you need to frame it in a particular TT/system context.

We know $5k can improve an analog rig a good amount, so is the upgrade in DSD ADC going to be an even larger amount of improvement for the same money?

Hard to say. I shared my opinion on that.

 

[Audiophile Bill]

Yes it makes an audible difference. Enough for me that it's the only way I will ever listen to DSD now with passive speakers. Unless going active where the DSP on the DSD is a must. Then you have to decide what the lesser of the 2 compromises are, passive crossovers, or multibit SDM DSP. In this situation, the lesser of the 2 comprimises will likely be the multibit SDM. After all it's still the finest way possible to do DSP. Far beyond what's possible on the best Sharc DSP chips.

As for the money spent on the analog source, I find it far more cost effective to buy master tape DSD rips, guys like Bruce did from NativeDSD and Analogue sounds, and get my vinyl rips for free from my buddy with the $100K plus rig.

Try for yourself with HQplayer. When you play a native DSD track and DirectSDM isn't checked, it gets processed through the multibit SDM engine. When checked, it bypasses the multibit SDM engine. In most SDM DAC chips, you can't bypass this engine, but you can on some, but only with DSD. Your GG doesn't have the SDM/SRC processing in it. the DSD section is just a simple low pass filter. So if you send it DSD trough HQplayer with SDMdirect enabled, it doesn't go through any multibit processing. This is the purest path possible. In this scenario the only SRC/SDM the audio sees is the initial 1 bit from the original ADC/DAW. So what this shows is there's no substitute for using the best 1 bit DSD SDM/SRC algorithms the first time around. This can only be done with the original ADC/DAW package. This is another reason DSD is better that is overlooked by the PCM advocates. You can never have a signal path this pure from start to finish with PCM, unless you use an inferior R2R DAC. But that technology is outdated now. The last R2R chip the BB 1704 was unveiled back in 1998. The days of SDM tech, not living up to R2R standards are in the past now. The advanced SDM algorithms in DAW like Pyramix and HQplayer will put any R2R to shame.

The comment about getting your rips with a mate with a $100k rig defeats the point of the question.

On the SDM thing - I concur since I use this mode with hqplayer and it is the best way to play dsd in my system.

 

[Blizzard]

So to reframe in my system context.

Turntable is Kuzma Ref with Triplanar VII and Lyra Kleos through BMC MCCI.

Option 1. Buy a Merging Horus and associated software for not sure price but let's just say $7k

Or

Optio. 2. Buy an Air Tight PC1 or Lyra Etna or whatever very high end cartridge. Use existing Tascam.

What recording sounds better? I reckon option 2.

Bruce would be a better guy to answer that. With the Merging HAPI, you will get an identical copy of the vinyl. So the bottleneck after that will be the analog source. And as Tbone mentioned earlier, the skill to make the rips is an essential part of the picture. I'd say if you can't tell your vinyl apart now from the digital with the Tascam, upgrade the analog instead.

No ADC on the planet will make your analog source any better.

 

[Audiophile Bill]

Bruce would be a better guy to answer that.

Bruce said to me on pm (before buying Tascam) that it was ~90% of Merging. So I guess would my analogue front end improve the final product by more than 10% moving to super high end cartridge?

 

[Bruce B]

Bruce said to me on pm (before buying Tascam) that it was ~90% of Merging. So I guess would my analogue front end improve the final product by more than 10% moving to super high end cartridge?

Definitely. I would certainly put money towards the front end. The Tascam has digital in/out, so you can always add a great ADC later. The better the TT, the quieter your rips would be. I've had to spend MUCH LESS time fixing pops/ticks/rumble in a higher end TT than a mediocre one. Besides, that last 10% on the digital side is where the point of diminishing returns kicks in. A Horus with ADC/DAC and Pyramix software is going to be close to $15k

 

[Blizzard]

Bruce said to me on pm (before buying Tascam) that it was ~90% of Merging. So I guess would my analogue front end improve the final product by more than 10% moving to super high end cartridge?

The best answer to this question can be answered by yourself. Can you hear any difference between your current vinyl rig, and the digital copies made from it? If this answer is no, then improving the analog will be the best choice.

 

[esldude]

Bruce said to me on pm (before buying Tascam) that it was ~90% of Merging. So I guess would my analogue front end improve the final product by more than 10% moving to super high end cartridge?

That advice from Bruce sounds quite believable. That last 10% improvement in most systems runs costs up by multiples with most types of gear. The performance of a basic good operating unit like the Tascam is going to be a huge price difference compared to the very best that gets you the last little improvement. As the Tascam already performs in absolute terms very well the 10% isn't that much compared to analog gear improvements that have more room for improving in the first place.

 

[Audiophile Bill]

That advice from Bruce sounds quite believable. That last 10% improvement in most systems runs costs up by multiples with most types of gear. The performance of a basic good operating unit like the Tascam is going to be a huge price difference compared to the very best that gets you the last little improvement. As the Tascam already performs in absolute terms very well the 10% isn't that much compared to analog gear improvements that have more room for improving in the first place.

Yeah - that was my working hypothesis too.

 

[Audiophile Bill]

The best answer to this question can be answered by yourself. Can you hear any difference between your current vinyl rig, and the digital copies made from it? If this answer is no, then improving the analog will be the best choice.

Not a lot in it if I'm honest. The rips don't have quite the same life and sparkle but they are damned close. They are all double dsd. Through the GG they are hugely enjoyable so happy with the final product.

 

[Audiophile Bill]

Definitely. I would certainly put money towards the front end. The Tascam has digital in/out, so you can always add a great ADC later. The better the TT, the quieter your rips would be. I've had to spend MUCH LESS time fixing pops/ticks/rumble in a higher end TT than a mediocre one. Besides, that last 10% on the digital side is where the point of diminishing returns kicks in. A Horus with ADC/DAC and Pyramix software is going to be close to $15k

Okay - thanks Bruce. That is good enough for me then. $15k actually improves a vinyl rig by a lot IMHO.