384khz?

[Qwin]

I'm no expert on digital, but there seems to be some confusion in this thread between the audio frequency and sample rate of the recording?
Sample rate being the number of measurements taken per second to try and capture any played frequency.
In order to record a smooth facsimile of the sound wave the sample rate has to be many times faster than the time span of the audio frequency wave.
Its like plotting a line graph of the wave, the more points you register the smoother the curve.

That's how I've understood it, bit rate v sample rate on the graph axis.

Or am I tilting at windmills

[Macca]

I'm no expert on digital, but there seems to be some confusion in this thread between the audio frequency and sample rate of the recording?
Sample rate being the number of measurements taken per second to try and capture any played frequency.
In order to record a smooth facsimile of the sound wave the sample rate has to be many times faster than the time span of the audio frequency wave.
Its like plotting a line graph of the wave, the more points you register the smoother the curve.

That's how I've understood it, bit rate v sample rate on the graph axis.

Or am I tilting at windmills

Increasing the sampling frequency increases the sound frequency you can record, The higher the sampling rate, the higher the frequency. That's all it does, increasing the sampling rate does not give a 'better look' at the frequencies lower down. The accuracy of the sampling can't get any higher.

[Rothchild]

I'm no expert on digital, but there seems to be some confusion in this thread between the audio frequency and sample rate of the recording?
Sample rate being the number of measurements taken per second to try and capture any played frequency.
In order to record a smooth facsimile of the sound wave the sample rate has to be many times faster than the time span of the audio frequency wave.
Its like plotting a line graph of the wave, the more points you register the smoother the curve.

That's how I've understood it, bit rate v sample rate on the graph axis.

Or am I tilting at windmills

No, you're right. The maximum audio frequency that can be reconstructed from a digital sample is half the sample rate. Hence 44.1kHz being the sample rate for CD, as for most people 20kHz is the upper end of the audible spectrum. So cd's can contain audio up to 20kHz and then there's a bit of space left for the roll-off of a low pass filter that stops anything over 22.05kHz being fed in to the converter.

One of the (more plausible) arguments for higher sample rates isn't that there's an awful lot more useful, audible, information above 20kHz, but that the necessary phase issues and 'ringing' that are caused by the low pass filter are moved up in to an area where they are unlikely to be heard.

[Rothchild]

The point I am making (or trying to) is that I am sure that improvements are possible, audible and measurable at some intermediate sampling frequency. Great if 96khz is achievable but why exclude the potential gains from some intermediate rate (say 48khz).

I do not claim to have golden ears but I have never found the high frequency performance of cd to be comparable to an equivalent (level) analogue system.

Geoff

Ahh, I get you.

Yeah, it's a fair point. One of the most interesting things to come out of the materials that came out around the release of MQA is the discussion of the optimum sample rate (which from memory was about 56kHz) allowing for the realistic most extreme top frequencies created by musical instruments (cymbals and the like) and allowing for a soft enough roll off of the LPF to prevent ringing, phase and aliasing getting in to the audible spectrum.

As I remember the argument continued that whilst 56kHz would be plenty sufficient it's non-standard, we have 96kHz as a standard sample rate and therefore that would do.

[Sherwood]

Ahh, I get you.

Yeah, it's a fair point. One of the most interesting things to come out of the materials that came out around the release of MQA is the discussion of the optimum sample rate (which from memory was about 56kHz) allowing for the realistic most extreme top frequencies created by musical instruments (cymbals and the like) and allowing for a soft enough roll off of the LPF to prevent ringing, phase and aliasing getting in to the audible spectrum.

As I remember the argument continued that whilst 56kHz would be plenty sufficient it's non-standard, we have 96kHz as a standard sample rate and therefore that would do.

That would make sense to me as 96/192 seemed rather excessive to capture these frequency outliers! It is entirely possible that the benefits I believe I hear from these recordings lie in the boundaries either side of the 20-20k range.

[Light Dependant Resistor]

I'd agree. But the cause of that is nothing to do with frequency response. Even the best RTR will only manage the same as CD. Vinyl in a best case scenario is about 24Khz, if you are using a moving coil cart, less with MM. Of course it still has to be on the recording to hear it, mics again, and even then it has to get past the low cut filter on the cutting head. In practice all 3 mediums are within a whisker of each other in terms of upper limit of response: logically, vinyl's subjective superiority in that area must be due to something else.

However a failed patent by Murray Crosby created the copying of that contained information (somehow) which created research into companding
Companding being https://en.wikipedia.org/wiki/Companding products then that were made by Dolby and DBX and many others enabled raw devices to
provide much greater dynamic range. DBX and Dolby have been a feature of almost every recording since 1965

Now why audiophiles have yet to catch up with the benefits companding provides remains a deep mystery, hopefully now to be investigated.
A cursory look provides very simple somewhat crude methods have been offered to consumers ie Dolby B. or Type 2 DBX simply because there is no discussion
or interest.

Surely this is what a forum like this is about, to raise awareness of an attribute for audio reproduction.

Anyone with a Dolby A or other more modern equivalents like Dolby SQ, hands up.

Back to Murray Crosbys invention ..RIAA is a reverse emphasis deemphasis curve, anyone know why ?

DBX too need to be contacted to start building a Type 4 digital domain product, and not hide behind
the fence offering Type 4 just with their poor analog stages. DBX have every ability to release
a world class product, holding a major slice of patents. need I say any more Harmon owners of DBX
now taken over by Samsung. Oh Dear !

Maybe as a suggestion Nebo should demonstrate companding equipment, by simply recording a CD
on to a audio hard disk recorder with and without companding. Could be interesting

[the_doc735]

so after the many technical arguments in this thread (which I can't understand), would there be any point in "upgrading" my audiolab M-DAC for a chord hugo? The 44KHz seems the kindest to my (tinnitus) ears, 192 is definitely out for my ears!! Seems obvious that high quality hifi sound is only available if it is recorded at the same high quality to begin with. I always have to reduce the treble in my software as it is far too sharp and sizzling! I personally do not think that I would notice any difference between the sound quality of these two DACs, but interesting to see anyway? Thanks for all the responses!

A few misconceptions to clear:
* Hugo is battery powered and cannot be powered any other way.
* Hugo charger imparts absolutely no noise back into the Hugo direct signal path, due to Hugo's isolation topology, so keeping the charger plugged in is fine. However, as with any ac wallwart it may backflow noise into the rest of a system so use care to isolate it as well as possible.
* The Hugo has undergone some chassis changes (larger holes for most inputs/outputs). If its new, anyone would be insured to get the revised one, of course.

[Rothchild]

would there be any point in "upgrading" my audiolab M-DAC for a chord hugo? The 44KHz seems the kindest to my (tinnitus) ears, 192 is definitely out for my ears!! Seems obvious that high quality hifi sound is only available if it is recorded at the same high quality to begin with. I always have to reduce the treble in my software as it is far too sharp and sizzling! I personally do not think that I would notice any difference between the sound quality of these two DACs, but interesting to see anyway? Thanks for all the responses!
.

Aside from bragging rights and a box swap fix it seems unlikely, if you're already turning the treble down in the audible spectrum you probably don't need any more info beyond that. Perhaps you should look at the rest of your system and see if you can't come up with something that has a naturally more pleasing sound to you?

Interestingly the spec for the m-dac says it's 20Hz-20kHz (despite being able to decode higher sample rate files) the Hugo just doesn't publish its analogue frequency response. It's a relevant point to consider I think, just because the chip can decode and put put audio signals up to 192kHz is the analogue stage up to it too (and not just the dac output, the amp input and the speakers too)?

[Rothchild]

However a failed patent by Murray Crosby created the copying of that contained information (somehow) which created research into companding
Companding being https://en.wikipedia.org/wiki/Companding products then that were made by Dolby and DBX and many others enabled raw devices to
provide much greater dynamic range. DBX and Dolby have been a feature of almost every recording since 1965

Now why audiophiles have yet to catch up with the benefits companding provides remains a deep mystery, hopefully now to be investigated.
A cursory look provides very simple somewhat crude methods have been offered to consumers ie Dolby B. or Type 2 DBX simply because there is no discussion
or interest.

Surely this is what a forum like this is about, to raise awareness of an attribute for audio reproduction.

Anyone with a Dolby A or other more modern equivalents like Dolby SQ, hands up.

Back to Murray Crosbys invention ..RIAA is a reverse emphasis deemphasis curve, anyone know why ?

DBX too need to be contacted to start building a Type 4 digital domain product, and not hide behind
the fence offering Type 4 just with their poor analog stages. DBX have every ability to release
a world class product, holding a major slice of patents. need I say any more Harmon owners of DBX
now taken over by Samsung. Oh Dear !

Maybe as a suggestion Nebo should demonstrate companding equipment, by simply recording a CD
on to a audio hard disk recorder with and without companding. Could be interesting

Compansion is a tool for maximising signal to noise ratios in systems with limited dynamic range, 24bit is not a limit on dynamic range (indeed it can capture sounds quieter than can be heard and louder than can be tolerated). Most discussion on the various boards here seems to show that folk want to 'extract' more of what's in the groove or 'get every last bit' from their source, aside from a bit of corrective eq there doesn't seem to be a lot of hunger for home remixing.

RIAA is a necessary evil to prevent the needle skipping out of the groove, nothing more. (it's also the reason why all these 'the groove of a record is an absolutely true analogue of the recorded signal' claims aren't quite right - it's an exact analogue subject to the phase issues of a filter (at both the input and the output - if they are truely identical then they cancel out, but in the analogue domain 'truely identical' is almost impossible).

[Macca]

16 bit for home replay is already overkill, CD was originally going to have 14 bit dynamic range. Listen to a 24/192 recording downsampled to 16/44.1 with no mastering - it will rip your tits off with dynamic range. I was actually afraid for the speakers.

With mastering the dynamic range is reduced to a level acceptable for home replay, often nowadays it is reduced too much, but that is a separate issue.

There really is no need to worry about dynamic range of digital systems, it is just more tilting at windmills. But punters like to see bigger numbers and so the marketing departments give them what they want.