[SimonPac]

Well gents, I just couldn't keep quiet. DAC's the same-no. I have Wolfson 24-192, Crystal 24-192 and BB PCM63P-K. set up as 44.1-20. PCM63P -K spec was state of the art laser trimmed R2R, used in many pro projects including digital desks, circa 2000. The K grade is trimmed to better than -96dBm THD which actually corresponds only to 16-bit resolution but executed absolutely perfectly. So ideal for CD. Even most top end commercial players tended to get only the basic grade of this chip; the goodies went to pro gear and people like Linn. They really are very good with the right filter. I have a PMD100 filter, actually phase-linear. I can switch instantly between any two of the DACs with the same CD 44.1 stream as source. The Crystal, a consumer level chipset, is quite 'musical' and 'forgiving', fairly sweet treble, lacks ultimate clarity and punch, fuzzy on bass instruments. Wolfson (it's not their top chip and I suspect the propriety filter used, have been getting round to doing an impulse test to check filter type..)..sounds polished, clean but ultimately a little muddled on CD. BB PCM63P-K with PMD100 filter I actually like best. Seriously, my player with the Crystal consumer DAC sounds a way away from the pro BB one. And that's all at 44.1-16. There's no way ever I'll agree they're the same. it will depend quite a bit on what you play them through, but high end headphones are amazing these days. The BB setup has slam, pace, clarity in complex passages and resolution after the others. IMO of course. Yes, I know they all measure THD-FR-PR-IMD-SNR-pitch- stability better than anything analogue except a very good amp.

Anyway....MAIN POINT

I have decided to try and find time to listen to the same file, analogue master, at 44.1-16 and 192-24, blind. I'll get someone else to randomly play them in order, either HD then CD or vice versa. I'll ask the person to record which was played first. I'll repeat this a few times and check and see how I did. How's that?

 

[bigshot]

Let us set the file up for you. We'll make it easy. No time at all. Just dropbox me the 24/192 file you want us to use and we'll do the rest.

 

[ev13wt]

I'll like to join in.

I propose foobars abx comparator. We could post the results here.

 

[bigshot]

PM me if you'd like to help.

 

[SimonPac]

I set up the HD 192-24 461 Ocean Boulevard file and re-sampled down using Audacity which I found easiest. I have to use the Cambridge USB DAC for replay which is not my favourite.

I'll see how I get on before I make a fool of myself! I still think there are meaningful if certainly not immediately blindingly obvious differences. I would concede you probably won't be able to detect them readily and reliably on a quick burst. I have one track looping between the two resolutions on at the moment.

I corresponded with my old acquaintance briefly today. He said he uses 44.1-24 for recording and mastering onto CD. I asked whether he had said 48-16 was transparent, and first he said 'nothing is totally transparent'. Then he said 'it's transparent enough'. If you're concerned I'm making this up I'll ask him if I can name him and his credentials.

 

[SimonPac]

BTW I was probably confusing direct to disc and direct metal mastering in places in my comment on vinyl at Abbey Rd. They seemed to do both as current work as of 2015 and of course the Mumford and Sons was DMM.

 

[gregorio]

DAC's the same-no.

No, you're misrepresenting what has been said. Modern DACs designed for high fidelity sound the same. There are some DACs which would not qualify, for example the Pono or potentially an old DAC.

G

 

[pinnahertz]

Well gents, I just couldn't keep quiet. DAC's the same-no. I have Wolfson 24-192, Crystal 24-192 and BB PCM63P-K. set up as 44.1-20. PCM63P -K spec was state of the art laser trimmed R2R, used in many pro projects including digital desks, circa 2000. The K grade is trimmed to better than -96dBm THD which actually corresponds only to 16-bit resolution but executed absolutely perfectly. So ideal for CD. Even most top end commercial players tended to get only the basic grade of this chip; the goodies went to pro gear and people like Linn. They really are very good with the right filter. I have a PMD100 filter, actually phase-linear. I can switch instantly between any two of the DACs with the same CD 44.1 stream as source. The Crystal, a consumer level chipset, is quite 'musical' and 'forgiving', fairly sweet treble, lacks ultimate clarity and punch, fuzzy on bass instruments. Wolfson (it's not their top chip and I suspect the propriety filter used, have been getting round to doing an impulse test to check filter type..)..sounds polished, clean but ultimately a little muddled on CD. BB PCM63P-K with PMD100 filter I actually like best. Seriously, my player with the Crystal consumer DAC sounds a way away from the pro BB one. And that's all at 44.1-16. There's no way ever I'll agree they're the same. it will depend quite a bit on what you play them through, but high end headphones are amazing these days. The BB setup has slam, pace, clarity in complex passages and resolution after the others. IMO of course. Yes, I know they all measure THD-FR-PR-IMD-SNR-pitch- stability better than anything analogue except a very good amp.

Interesting opinions...but not actually valid for real evaluation. I'll elaborate if you don't get why.

Anyway....MAIN POINT

I have decided to try and find time to listen to the same file, analogue master, at 44.1-16 and 192-24, blind. I'll get someone else to randomly play them in order, either HD then CD or vice versa. I'll ask the person to record which was played first. I'll repeat this a few times and check and see how I did. How's that?

Not good enough. You need to get the person out of the equation, verify how each file was generated (otherwise you're testing down-sampling, etc..also) and perform way, way more trials, ABX, with randomized X.

 

[pinnahertz]

I set up the HD 192-24 461 Ocean Boulevard file and re-sampled down using Audacity which I found easiest. I have to use the Cambridge USB DAC for replay which is not my favourite.

I'll see how I get on before I make a fool of myself! I still think there are meaningful if certainly not immediately blindingly obvious differences. I would concede you probably won't be able to detect them readily and reliably on a quick burst. I have one track looping between the two resolutions on at the moment.

Just as long as you manage to eventually get to real DBT/ABX, level matched, and do enough trials. Realize that your results will be specific to that particular hardware, not indicative of comparing the quantization depth and frequency alone.

I corresponded with my old acquaintance briefly today. He said he uses 44.1-24 for recording and mastering onto CD. I asked whether he had said 48-16 was transparent, and first he said 'nothing is totally transparent'. Then he said 'it's transparent enough'. If you're concerned I'm making this up I'll ask him if I can name him and his credentials.

I'm not clear why, given all the back/forth so far, you haven't already named him, but it doesn't really matter, it's just another opinion, and no doubt based on fully sighted and biased subjective impression. At this point I'd probably say don't embarrass him.

 

[bigshot]

I'll see how I get on before I make a fool of myself!

To me, I feel like a fool when I believe things that aren't true. And I feel like a bigger fool if I resist doing what I need to do the uncover the truth.

I've offered to set you up a controlled test that is easy for you. All you have to do is listen and rank samples. If you want to know the truth, that's the easiest way to find out. If you do more sighted, uncontrolled, casual comparisons, you're not doing what it takes to find out, and we'll just keep saying that you need to apply controls before we'll listen to your opinions.

 

[pinnahertz]

With regard to using Foobar and the ABX component to do some testing...

Lest I come off as an even bigger curmudgeon than I already am, I'll preface by saying I'm a pretty hard-core ABX guy. I own a hardware ABX comparator, and have designed and administrated many tests over several decates.

Foobar has several issues to be aware of. There is a strong tendency to make rapid comparisons followed by rapid decisions. This gives rise to the "ABX=High Pressure" objections. In reality, there is no time limit, and tests could be conducted over weeks or months or more. There is also a tendency to take too few trials. This results in rather high "noise" in the data. To improve the data quality somewhere between 12 and 16 trials is good for a single file pair, and there should be at least a few file pairs. There should also be a file pair that are exactly identical as a control.

However, the big problem with Foobar and ABX is that cheating is very, very easy. And it's highly motivated by the desire to perform well. Files may be named ambiguously, but their size will definitely be a strong indicator. It's also very easy to grab more information from any file and capture all one would need to form a strong bias. This is the real problem with an non-proctored test of this kind, as cheating adds a very high "noise" component to the composite data. There is no good way to control this, unfortunately.

It's also worth noting that the comparison ultimately includes the method of down/upsampling used to create the files, as well as the ability of the DAC in use to handle all bit rates/depths equally well. These are also data noise sources, so to keep them low there is benefit to using many different DACs, headphones, speakers, etc. But that also means there would need to be a LOT of data collected to be really meaningful.

I'm not saying "don't do it", I'm just raising the flags now before conclusions are drawn by individuals with "noisy" data, and very little of even that. Remember, the process is not evaluating an individual's abilities, it's using the individual's senses in a series of subjective comparisons along with statistical analysis to develop a means of objective measurement. There literally is no need to "score well". There is a strong desirability to report findings honestly.

 

[danadam]

Files may be named ambiguously, but their size will definitely be a strong indicator. It's also very easy to grab more information from any file and capture all one would need to form a strong bias.

This one I don't understand. Yes, you cn infer from the files if they are different or not, but isn't the whole point of ABX plugin that when X plays you don't know which file plays? How does the access to the files help you?

 

[bigshot]

I always put all the samples in a single file that is the same as the highest bit/sampling rate sample. It's easier to be sure it will all be fair when you make it hard not to be honest.

 

[SimonPac]

I can only present results with one DAC and it's a fairly cheap one, a CA DACMagic 100. I'm at the mercy of s/w rate change of course. I intend to get my wife or daughter to randomly choose a file and note it, I've just added -Hi and -Lo at the end of the titles so they can do that.

 

[SimonPac]

I can hide the rate lamps easily enough. This will take a while because I'm aware the differences (if any) are fairly subtle. My memory is that they are fairly clear if you quietly listen to a whole album. The threshold for detection with a quick comparison with most gear is probably below 44.1-16 and not above it, agreed.