MQA spectrum plots

[mansr]

I see a business opportunity here for an enterprising digital filter designer to make plugins that finish the "unfolding" and "de-blurring" for software decoders such as Roon. Given that the Blusound decoder was designed with the NAD PWM DAC architecture, it should already have that transfer function considered into it's final (apparently upsampling) filter design. I'd pay for that to use with my M51 DAC out of HQP from Roon. Just saying. Anyone?

 

I mean, even if MQA is a botch, I'd rather hear it fully decoded to hardware spec accoring to their purported end-to-end strategy.

 

Good thread. Keep the reverse-engineering going. It's only a matter of time before it's cracked.

Be patient. It's a lot of code.

[mansr]

This is a grovelling apology. I made a mistake running the decoder. The results above are not accurate.

 

Having corrected the mistake, the output of the decoder is looking much better. I'll post graphs soon.

[jtwrace]

This is a grovelling apology. I made a mistake running the decoder. The results above are not accurate.

 

Having corrected the mistake, the output of the decoder is looking much better. I'll post graphs soon.

 

Might want to ask Chris to delete that post to avoid any confusion.

[sdolezalek]

Thank you for what couldn't have been an easy apology. I'd agree with jtwrace, but go one step further: why not have Chris delete this entire thread and just start a new one when you have you have the new graphs ready.

[crenca]

Thank you for what couldn't have been an easy apology. I'd agree with jtwrace, but go one step further: why not have Chris delete this entire thread and just start a new one when you have you have the new graphs ready.

 

I would recommend amending the thread and not deleting it - if someone else is getting similar results they can learn from the mistake...

[bibo01]

This is a grovelling apology. I made a mistake running the decoder. The results above are not accurate.

 

Having corrected the mistake, the output of the decoder is looking much better. I'll post graphs soon.

 

All graphs are wrong?!

[audiventory]

Might want to ask Chris to delete that post to avoid any confusion.

 

Me seems, graph need stay. Otherwise, we lost knowledges about the decoder limitations.

[Nikhil]

Might want to ask Chris to delete that post to avoid any confusion.

 

No need to delete the original posts - a simple note on there with a link to the revised graphs would be sufficient.

There's no shame in owning up to genuine discovery given the nature of the MQA discussion.

 

This thread has been one of the best in a long time on here - genuine open technical discussion.

[jtwrace]

No need to delete the original posts - a simple note on there with a link to the revised graphs would be sufficient.

There's no shame in owning up to genuine discovery given the nature of the MQA discussion.

 

This thread has been one of the best in a long time on here - genuine open technical discussion.

 

It's not about the shame (there is none) it's about wrong data being presented. That's all. No data is better than bad data. It's that simple. It causes confusion.

[mmerrill99]

It's not about the shame (there is none) it's about wrong data being presented. That's all. No data is better than bad data. It's that simple. It causes confusion.

+1,000 Completely agree

 

What exactly was the point about "This thread has been one of the best in a long time on here - genuine open technical discussion."

 

What exactly can be learned from technical discussion about wrong data - beware of scientism here!

[lmitche]

+1,000 Completely agree

 

What exactly was the point about "This thread has been one of the best in a long time on here - genuine open technical discussion."

 

What exactly can be learned from technical discussion about wrong data - beware of scientism here!

What is the meaning of scientism here?

[mmerrill99]

What is the meaning of scientism here?

Probably misconstrued it but it's the closest term I could find to what I mean - The idea that any test (even if flawed) makes for valid scientific investigation.

 

Fair play to mansr for his honesty - many would try to bluff it out rather than admit they had made a mistake - there's absolutely no shame in this, in fact it's refreshing to see. I wondered what value was served in allowing to stand 5 pages of wrong data & technical discussions of this wrong data?

[Vincent1234]

Probably misconstrued it but it's the closest term I could find to what I mean - The idea that any test (even if flawed) makes for valid scientific investigation.

 

Fair play to mansr for his honesty - many would try to bluff it out rather than admit they had made a mistake - there's absolutely no shame in this, in fact it's refreshing to see. I wondered what value was served in allowing to stand 5 pages of wrong data & technical discussions of this wrong data?

A quick, corrected version by Mansr would certainly help to clear all the wrong assumptions being made in this thread..

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

[Jud]

Probably misconstrued it but it's the closest term I could find to what I mean - The idea that any test (even if flawed) makes for valid scientific investigation.

 

Fair play to mansr for his honesty - many would try to bluff it out rather than admit they had made a mistake - there's absolutely no shame in this, in fact it's refreshing to see. I wondered what value was served in allowing to stand 5 pages of wrong data & technical discussions of this wrong data?

 

+1 regarding refreshing. I have a feeling a lot of hard, irrelevant words could be avoided and discussions could be consistently more valuable if we all tried to follow this example. Instead of "playing defense:" if you're wrong, state it plain, apologize, and move on; if it's just your opinion, say so, allow the other person his/her opinion and rejoice in the variety of viewpoints we're exposed to.

[realhifi]

I'm not following. Are the graphs already posted here not correct? Any of them? Is this from a certified lab? Hobbiest?

[Vincent1234]

I'm not following. Are the graphs already posted here not correct? Any of them? Is this from a certified lab? Hobbiest?

 

Yes, they are all wrong.

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

[mansr]

Yes, they are all wrong.

 

They're not wrong per se, they're just not showing what I thought they were. I'm starting a new thread shortly where we can begin afresh.

[Vincent1234]

They're not wrong per se, they're just not showing what I thought they were. I'm starting a new thread shortly where we can begin afresh.

Thanks, we are looking forward to that.

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

[Tony Lauck]

Yes, lowering the level by 3 dB leaves only the fundamental and the alias frequency one would expect from the frequency response plot. Although an actual recording typically has some headroom, I see no reason for the filter to break down like this with a full-range input.

 

Filters can create peaks that are larger than the input waveform. (This is obvious if they do boosts in certain frequency ranges, but it can also happen with cuts, e.g. square wave filtered to its fundamental has peaks 4/ pi larger than original peaks.) So a filter in a DAC must be designed to have a certain amount of headroom. This has to be done by reducing the input signal amplitude, and this means that there will be a lower signal to noise ratio out of the DAC chip. Designers are given a tradeoff: more distortion on some high level signals vs. lower noise on low level signals. This is one of the classic design tradeoffs that are market oriented: high end customers will care about the distortion, but probably not a low level of measured noise. Mid-fi customers will work off of specifications and pick the distorting DAC because it measures better. (This is only because the published measurements didn't use the proper test signals.)

 

These issues are readily audible if one has a computer audio system and can perform digital and analog volume control at different points in the playback chain. Unfortunately, this is impossible with MQA encoded signals because of the proprietary nature of the encoding and decoding process.

[mav52]

mansr , I look forward to you're new plots. I learned a lot from the old ones.

[mansr]

New thread: http://www.computeraudiophile.com/f8-general-forum/mqa-technical-analysis-31311/