You are here

A-D & D-A converters, budget vs high-end

All about the tools and techniques involved in capturing sound, in the studio or on location.

Re: A-D & D-A converters, budget vs high-end

Postby Hugh Robjohns » Wed Aug 21, 2019 3:19 pm

Equipment testing is a fascinating subject. We can certainly measure a lot of things that we can't ~apparently~ hear, and there may well be some things that we can hear but can't measure (reliably).

The well known pro-audio author Dr John Watkinson has suggested that, as the more advanced lossy data-codecs are based around a model of human hearing, they could actually form the basis of a more accurate means of assessing the quality/accuracy of audio equipment, sine they could reflect the things that we actually hear. I believe he was specifically considering loudspeakers when he made this suggestion.

The idea would be to play a very high quality source recording through a lossy codec, and to reduce the coding bit rate step by step until artefacts became audible. The better the loudspeaker (or other equipment), the sooner the artefacts would be revealed... and vice versa.

As for the lost of converters in your link, I usually mention them in my reviews because I get asked to when I meet people at trade shows etc -- just as I'm asked for photos of the insides of the products. And these things can be revealing and informative.

For example, knowing which converter chips has been used gives an indication of the cost -- and therefore quality -- of components used, and if a manufacturer is only prepared to employ a low-cost converter they've probably used low cost/quality components elsewhere too.

Moreover, if the specific converter chip manufacturer claims it has, say, a 118dB dynamic range but the equipment only measures with a 112dB range, that suggests the design/build is compromised in some way as the performance of the converter is not being realised. So what else hasn't been designed very well (it's usually the power supply or the clocking... :-) )

H
User avatar
Hugh Robjohns
Moderator
Posts: 24350
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: A-D & D-A converters, budget vs high-end

Postby Ariosto » Wed Aug 21, 2019 3:37 pm

blinddrew wrote:I'm just wondering how long it takes to back up a reasonable sized project to the cloud at that kind of sample rate? Especially if you're doing the whole project not just mixes and stems.

It does take a long time. Uploading is so slow in comparison to downloading. I have 80-100 MBPS download but only 6MBPS upload. For this reason I never upload complete projects as these are often 1, 2 or 3+ Gigabytes. So I only upload the rendered .wav files but these can be well in excess of 500GB each at 96K sample rate. Two or three files and it amounts to best part of an hour to upload.
Ariosto
Frequent Poster
Posts: 805
Joined: Sun May 04, 2008 12:00 am
Location: LONDON, UK

Re: A-D & D-A converters, budget vs high-end

Postby Hugh Robjohns » Wed Aug 21, 2019 3:46 pm

Ariosto wrote:Uploading is so slow in comparison to downloading. I have 80-100 MBPS download but only 6MBPS upload.

If you want faster uploads just talk to your broadband supplier and I'm sure they could assist for a small increase in your monthly fee.... I have 66Mbps down and 18Mbps up here.

H
User avatar
Hugh Robjohns
Moderator
Posts: 24350
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: A-D & D-A converters, budget vs high-end

Postby ef37a » Wed Aug 21, 2019 4:30 pm

Hugh Robjohns wrote:
Ariosto wrote:Uploading is so slow in comparison to downloading. I have 80-100 MBPS download but only 6MBPS upload.

If you want faster uploads just talk to your broadband supplier and I'm sure they could assist for a small increase in your monthly fee.... I have 66Mbps down and 18Mbps up here.

H

My download is 20-24M wifi (laptop 1mtr from router) but is 37M on copper. The upload is a very consistent 9M wifi or copper, anyone know why?

I am with Talk Talk who are ******es to deal with but in over 4 years I have have virtually no breaks in service or slow downs.

I am FTTC and the cab is but 50mtrs down the frog. My 9M upload is ten times better than it was on direct landline. Perhaps you would have to go fibre to client to better 9M?

BTW, serves me right trying to be a smart****! I was meaning the petabytes of data the astrophycisists handle!

Dave.
ef37a
Jedi Poster
Posts: 10601
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk

Re: A-D & D-A converters, budget vs high-end

Postby Aural Reject » Wed Aug 21, 2019 4:50 pm

Standard fibre used to be “40/10” so max 40 down and 10 up. The ‘super fast” stuff was sold at 80/20.

The download speed is amongst other things down to the distance between the cabinet and the router and the quality of the cabling and connections.

The upload is usually tweakable server side and is limited by the ISP so I’d imagine the 9 up is easily achievable by wired and wireless methods (download over WiFi is usually slower than it is over cable).

I’ve got Virgin 350 here....and I actually get 320 ish down over WiFi, 360 cabled and about 26 up over both. It ain’t that cheap though.
User avatar
Aural Reject
Frequent Poster
Posts: 899
Joined: Fri May 02, 2003 12:00 am
Location: Lancashire born, living in Yorkshire :s

Re: A-D & D-A converters, budget vs high-end

Postby Hugh Robjohns » Wed Aug 21, 2019 4:57 pm

ef37a wrote:My download is 20-24M wifi (laptop 1mtr from router) but is 37M on copper. The upload is a very consistent 9M wifi or copper, anyone know why?

If you're asking why the wifi link is slower than a cabled connection, it's because it's wireless! ;-) It's quite normal for a wired connection to be faster than wifi -- although the differences are getting smaller with the latest high-speed connection standards.

There are lots of factors affecting wifi speed. Firstly, there's the specific connection standard your laptop and router are using (802.11b, 802.11g, 802.11n, 802.11ac.. etc). In general, the older standards are slower. Then there's the router channel and frequency you're using and whether neighbours are sharing the same channel(s) -- interference will slow things down dramatically. And then there's the distance between laptop and wifi router and the number of other devices sitting on your wifi channel at the same time... More of both being slower, of course.

If you have a smart phone you can download various apps that indicate the local channel occupation and signal strengths, which can be very helpful in optimising your wifi channel selection and router placement.

Perhaps you would have to go fibre to client to better 9M?

No. I'm, FTTC as well and I have double that upload speed... (and the cabinet is further away than yours). I suspect the 9Mbps limit is imposed by the provider to manage data bandwidths, rather than being a physical or hardware limitation.
User avatar
Hugh Robjohns
Moderator
Posts: 24350
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: A-D & D-A converters, budget vs high-end

Postby ef37a » Wed Aug 21, 2019 6:15 pm

Hugh, sorry I asked wrongly. It is the consistent 9M (9.2 as a rule) that I was boggled by. I can see wif fi being slower than copper.

The laptop is a rather elderly HP i3 g6 and perhaps a better USB network dongle would help? Not that I am all that bothered, 22M is plenty good enough for my needs and the 9M up a great boon.
I shall have a much later W10 laptop by Christmas anyway, might be faster?

Dave.
ef37a
Jedi Poster
Posts: 10601
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk

Re: A-D & D-A converters, budget vs high-end

Postby Hugh Robjohns » Wed Aug 21, 2019 6:22 pm

ef37a wrote:Hugh, sorry I asked wrongly. It is the consistent 9M (9.2 as a rule) that I was boggled by.

Ah... well that's just a network management thing -- an arbitrary soft limit imposed by the network providers to manage data throughput.

H
User avatar
Hugh Robjohns
Moderator
Posts: 24350
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: A-D & D-A converters, budget vs high-end

Postby Tim Gillett » Wed Aug 21, 2019 7:57 pm

Jack Ruston wrote:
Tim Gillett wrote:
Jack Ruston wrote:
hobbyist wrote:How do you measure how 'quality' <<improves (y)our work>> when the vast majority of consumers cant tell and dont care ?

I don't know how many more ways I can say it...The conversion changes the decisions I make - The eq points are different, the compression time constants, reverbs etc. The balance changes. Hugh has explained this too.
Jack, there must be reasons why the different converter changes the eq points, compression time constants, reverbs etc. What would they be? How does it relate to the performance specs of the respective converters?

That's the big question. I don't know. And I don't know if they know, but that they just can't do it at the price point, or if they don't know, and just continue to champion frequency response at 20k, and noise floor, as a way of saying 'look, it's perfect'.

For me that different high end converters would audibly change the eq points or compression time constants or reverb seems highly unlikely. And if there were audible changes I wouldnt expect them to be described in those ways. Hugh what's your take on Jack's claims?
Tim Gillett
Frequent Poster
Posts: 1784
Joined: Wed Jan 30, 2013 1:00 am
Location: Perth, Western Australia

Re: A-D & D-A converters, budget vs high-end

Postby Jack Ruston » Wed Aug 21, 2019 8:10 pm

Oh yeah, now I come to think of it, you're absolutely right. Thank goodness you were here to help me see how wrong I've been. :)
Jack Ruston
Frequent Poster (Level2)
Posts: 3769
Joined: Wed Dec 21, 2005 1:00 am

Re: A-D & D-A converters, budget vs high-end

Postby James Perrett » Wed Aug 21, 2019 8:50 pm

Tim Gillett wrote:
For me that different high end converters would audibly change the eq points or compression time constants or reverb seems highly unlikely.

Tim - I think you've misunderstood what Jack is saying. His choice of reverb, eq or compression parameters would be affected by the characteristics of the convertor. He's not saying that the convertor would appear to change these parameters themselves. Now I don't know which particular convertors Jack has experienced this with but I can certainly believe that small changes in something like jitter characteristics or filtering could cause this.

The biggest problem in convincing the sceptics is the need for extremely good monitoring to hear these changes reliably. If you don't have good monitoring then I can quite understand you wondering what all the fuss is about.
User avatar
James Perrett
Moderator
Posts: 8361
Joined: Mon Sep 10, 2001 12:00 am
Location: The wilds of Hampshire
JRP Music - Audio Mastering and Restoration. JRP Music Facebook Page

Re: A-D & D-A converters, budget vs high-end

Postby Hugh Robjohns » Wed Aug 21, 2019 8:53 pm

Tim Gillett wrote:Hugh what's your take on Jack's claims?

My take is that I'd be wasting my time writing this reply... But here goes anyway...

It seems to me completely illogical for anyone without relevant experience to doubt those who do. Anyway, I've already described the benefits I associate with high-end gear, which are very similar to those Jack has described.

It's just like the aural equivalent of looking through a cleaner window -- you can see/hear things a little more clearly, and that makes it easier to make decisions about what processing is needed, and to apply it more delicately.

These are very small benefits, for sure, but when you're doing this stuff all day everyday, it's noticeable and it makes a real difference to the way you work... including having more finesse in the choices of things like dynamic time constants, EQ bandwidths and gains, reverb parameters and so on. And yes, many experienced clients notice and appreciate it too.
User avatar
Hugh Robjohns
Moderator
Posts: 24350
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: A-D & D-A converters, budget vs high-end

Postby hobbyist » Wed Aug 21, 2019 9:21 pm

Hugh Robjohns wrote:
Tim Gillett wrote:Hugh what's your take on Jack's claims?

My take is that I'd be wasting my time writing this reply... But here goes anyway...

It seems to me completely illogical for anyone without relevant experience to doubt those who do. Anyway, I've already described the benefits I associate with high-end gear, which are very similar to those Jack has described.

It's just like the aural equivalent of looking through a cleaner window -- you can see/hear things a little more clearly, and that makes it easier to make decisions about what processing is needed, and to apply it more delicately.

These are very small benefits, for sure, but when you're doing this stuff all day everyday, it's noticeable and it makes a real difference to the way you work... including having more finesse in the choices of things like dynamic time constants, EQ bandwidths and gains, reverb parameters and so on. And yes, many experienced clients notice and appreciate it too.

And that would be a logical fallacy called appeal to authority.

No one disputes it makes a difference. Only that you can not say one pane of glass is better than another when you come to the higher half of all glass panes.
hobbyist
Regular
Posts: 294
Joined: Mon Jul 01, 2019 12:52 am

Re: A-D & D-A converters, budget vs high-end

Postby Logarhythm » Wed Aug 21, 2019 10:12 pm

hobbyist wrote:No one disputes it makes a difference. Only that you can not say one pane of glass is better than another when you come to the higher half of all glass panes.
Um, I think you probably can. Indeed it is probably actually much easier to compare panes of glass than a/d converters, considering solely the property of transparency that Hugh mentioned - simply point a monochromatic beam at it, sweep across wavelengths of 400-800nm, and the "best" glass is surely the one for which the absorption value integrated across the measured range is lowest.
And yes, as someone who spent more hours than I care to remember locked in a very dark room taking optical measurements of various nanoscale structures constructed on glass slides, I can assure you that there are applications for which this genuinely does make a difference. Admittedly probably not for the average domestic double glazing, but then this is all becoming curiously analogous to the difference between the a/d requirements of a top end studio and engineer, as compared to those of amateurs like me ;)
Logarhythm
Frequent Poster
Posts: 764
Joined: Fri Mar 30, 2012 12:00 am
Location: On an intergalactic voyage to collect fresh electrons

Re: A-D & D-A converters, budget vs high-end

Postby hobbyist » Thu Aug 22, 2019 12:53 am

Logarhythm wrote:
hobbyist wrote:No one disputes it makes a difference. Only that you can not say one pane of glass is better than another when you come to the higher half of all glass panes.
Um, I think you probably can. Indeed it is probably actually much easier to compare panes of glass than a/d converters, considering solely the property of transparency that Hugh mentioned - simply point a monochromatic beam at it, sweep across wavelengths of 400-800nm, and the "best" glass is surely the one for which the absorption value integrated across the measured range is lowest.
And yes, as someone who spent more hours than I care to remember locked in a very dark room taking optical measurements of various nanoscale structures constructed on glass slides, I can assure you that there are applications for which this genuinely does make a difference. Admittedly probably not for the average domestic double glazing, but then this is all becoming curiously analogous to the difference between the a/d requirements of a top end studio and engineer, as compared to those of amateurs like me ;)

There are many other parameters to consider and that would say you cannot logically say which is best.

Your surely best is your opinion. I might prefer some uv filtering,
others may care about strength against breaking, and if your application makes a difference then that would make your choice different but not prove that any pane was THE best.
hobbyist
Regular
Posts: 294
Joined: Mon Jul 01, 2019 12:52 am

Re: A-D & D-A converters, budget vs high-end

Postby Terrible.dee » Thu Aug 22, 2019 1:37 am

Elephone wrote:Thanks... but what is the actual observable difference though? Is it a type of distortion or is it low background noise, or both?

I mean, would it more important to use high-end converters for recording music with quieter passages, like say a solo violin piece (or if you want to record a pin drop) ...or is the result better in terms of fidelity to the actual programme waveforms (noise aside) or both?

Thanks.

WELL DONE!

You just asked the Emporer why he is naked.

All you are going to get is highly vague, subjective answers, that really aren't answers at all.

Sure, you can tell a "Difference" but that difference is not a standard objective quality, that becomes more pleasing as you spend more cash.

I KNOW I like my Audient converters (Which are Burr Brows) better than the Focusrite ones, but as I describe what I like better, it seems I am describing the results of an analog path I prefer, not better ADDA converters. I probably like the preamps better, because when I skip the pres and go straight to the converters from the insert, then I'm not so sure anymore, I'm still pretty sure I like the Audient (Burr Browns) better...but now it's "elusive"

Also, if you need to pull out a scope, to "Proove" it's better, then it isn't better, your slight bump at 90 hz doesn't impress me if it needs to be SEEN to make its self known.

All in all, I believe AD/DA converters to be one of many scams being pulled in all areas of business, where prestige and ego are skillfully tied to factors that do little but stroke the aforementioned locals.

Yes, converters DO improve here and there over the years, but not nearly as often as the industry tells you they do, they are also a CHEAP CHIP. If a new converter hits the market, and it makes things sound better, it won't be long till EVERYONE is using it, or a copy.

For that reason, it may be expedient to make sure you use somewhat "Up to date" converters, but blowing money on "Top of the line" AD/DA is money down the toilet. The analog path is where the money is....My advice: Forget converters, keep them up to date and invest in a great analog path too and from.
Terrible.dee
Poster
Posts: 39
Joined: Tue Apr 18, 2017 9:12 pm

Re: A-D & D-A converters, budget vs high-end

Postby hobbyist » Thu Aug 22, 2019 3:46 am

Terrible.dee wrote:
Elephone wrote:Thanks... but what is the actual observable difference though? Is it a type of distortion or is it low background noise, or both?

I mean, would it more important to use high-end converters for recording music with quieter passages, like say a solo violin piece (or if you want to record a pin drop) ...or is the result better in terms of fidelity to the actual programme waveforms (noise aside) or both?

Thanks.

WELL DONE!

You just asked the Emporer why he is naked.

All you are going to get is highly vague, subjective answers, that really aren't answers at all.

Sure, you can tell a "Difference" but that difference is not a standard objective quality, that becomes more pleasing as you spend more cash.

I KNOW I like my Audient converters (Which are Burr Brows) better than the Focusrite ones, but as I describe what I like better, it seems I am describing the results of an analog path I prefer, not better ADDA converters. I probably like the preamps better, because when I skip the pres and go straight to the converters from the insert, then I'm not so sure anymore, I'm still pretty sure I like the Audient (Burr Browns) better...but now it's "elusive"

Also, if you need to pull out a scope, to "Proove" it's better, then it isn't better, your slight bump at 90 hz doesn't impress me if it needs to be SEEN to make its self known.

All in all, I believe AD/DA converters to be one of many scams being pulled in all areas of business, where prestige and ego are skillfully tied to factors that do little but stroke the aforementioned locals.

Yes, converters DO improve here and there over the years, but not nearly as often as the industry tells you they do, they are also a CHEAP CHIP. If a new converter hits the market, and it makes things sound better, it won't be long till EVERYONE is using it, or a copy.

For that reason, it may be expedient to make sure you use somewhat "Up to date" converters, but blowing money on "Top of the line" AD/DA is money down the toilet. The analog path is where the money is....My advice: Forget converters, keep them up to date and invest in a great analog path too and from.

Thank you.

At least I think you agreed with me.
hobbyist
Regular
Posts: 294
Joined: Mon Jul 01, 2019 12:52 am

Re: A-D & D-A converters, budget vs high-end

Postby Jack Ruston » Thu Aug 22, 2019 5:45 am

Spend money on the analogue stage...yes, like the analogue stage in your...converter.

Incidentally my converter uses a rather unusual chip, but your point that a lot are the same is absolutely right. But I'd suggest that there's a whole lot of stuff aside from the actual chip that makes the real difference to the sound of the box.
Jack Ruston
Frequent Poster (Level2)
Posts: 3769
Joined: Wed Dec 21, 2005 1:00 am

Re: A-D & D-A converters, budget vs high-end

Postby Wonks » Thu Aug 22, 2019 6:40 am

Indeed. Take the same A/D chip, put it in two different analogue paths within the converter package and the end results will be different. An awful lot of the 'sound' depends upon the support circuitry, not the A/D chip itself.
User avatar
Wonks
Jedi Poster
Posts: 9234
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: A-D & D-A converters, budget vs high-end

Postby CS70 » Thu Aug 22, 2019 8:16 am

Hugh Robjohns wrote:The well known pro-audio author Dr John Watkinson has suggested that, as the more advanced lossy data-codecs are based around a model of human hearing, they could actually form the basis of a more accurate means of assessing the quality/accuracy of audio equipment, sine they could reflect the things that we actually hear. I believe he was specifically considering loudspeakers when he made this suggestion.

Yes! It definitely would be possible to train some sort of classifier that way - there is enough consensus og what is "sound" and "artifact" that it would be well worth to try and see what happens.

Just thinking loud, if we hand-built a database of a few hundred lossy-encoded data files (say mp3s) classified into "good", "mediocre" or "bad" (there would be no problem to use the same song but encoded at different rates) we wouldn't really need any analytical model - the classifier would extract the relevant statistical properties during the training, based on the given classification. A key would be to find the relevant classes ("no artifacts", "good but with that type of artifact" etc) but the nice thing is these classes could be inferred by interviewing a sample of people who can recognize the differences.

Building and training such a classifier should be relatively straightforward. It could be possible to start with simple classes and just a few tracks, say ten, and see how the classifier does after being trained.

It could even be possible with unsupervised learning, but that's a bit tougher. Many years ago, as a pre-graduate student, I worked in a virtual group which used unsupervised-learning neural networks (a resurgent idea around 1994) to suggest new bands to people after they had listed the ones they liked - which is run of the mill now but back then was unheard of. Little we knew that Amazon would use the same idea a few years to great commercial success! :)
User avatar
CS70
Jedi Poster
Posts: 4082
Joined: Mon Nov 26, 2012 1:00 am
Location: Oslo, Norway
Silver Spoon - Check out our latest video  and the FB page

Previous

Who is online

Users browsing this forum: No registered users