The new look hifi+

I've actually made the (minimal) effort required to read most of the current issue and have to say I'm very pleased. Pretty much all of my previous criticisms have been addressed; I may actually start buying it again now.

 

Save me from buying it then...

 

Anyone who likes HMHB is OK in my book.

 

Perhaps..............

They didn't feature in my top three (not in the bottom three either mind you).

 

the old hifi+ was badly written, and looked and smelt great. The new HF+ is badly written and looks shit.

 

Is the next instalment of the Gregory/Quantum/Acuity we-can-measure-our-foo in there?

 

There is a Vertex-AQ advert in the latest issue presenting a graph purporting to show measured differences as a result of this 'research'. Of course it's completely impossible to interpret because there's no context, scales or anything that could possibly be used to make a judgement. (I guess that might be dangerous to the efficacy of the advert ). It just seems to be more of the same outrageous claims, but now with a mysterious graph to justify it.

So business as usual then.

 

I'm unfamiliar with the Vertex AQ advert.

Regular readers will remember 'Russ Andrews' being looked at by Advertising Standards in answer to a customer complaint. ASC did not find in Russ Andrew's favour.

If anyone is unhappy with claims made in advertising, then the process is there (and really easy) to register a complaint.

I simply mention this because there seems a view of general 'lack of accuracy' in advertising being discussed in the thread.

Not specifically targeting this comment at anyone

 

I have a lot of sympathy with the editors because it seems people actually want to read that stuff.
The tweaky enthusiast end of the hi-fi business has much in common with the cosmetics industry.
Hordes of middle aged wrinkly women (and increasingly men) spending millions on fancy creams and potions in the everlasting quest for the miracle beauty treatment. Pick up a lifestyle magazine and take a flick.....
Same thing happens in audio with millions spent on tweakery in an effort to squeeze just that little extra from the system. No understanding or real evidence required!

So what is the magazine editor to do?
In Alan's defence he has at least attempted some balance and it was brave to include the 'no foo' articles from the PMC chap.

 

Well, it's an advert, so viewer discretion is advised. I looked at the first graph, then the second, then the first, then the second and couldn't see any difference. Then I spotted it. Whatever it may mean.

If the research manages to explain some of the more contentious issues, then great, but it needs to be presented properly. If the research is not fit for wide publication yet, then I don't consider it fit for use in an advert as justification for a product's performance.

As far as the mag. goes, I think Alan's one of the best hifi journalists and have always enjoyed his writings. He's definitely more 'grounded' than the previous incumbent.

 

Your dreams are coming true:

http://www.vertexaq.com/images/stories/pdfs/ka paper - feb 2011 v 1.3 vertex only.pdf

 

I've tried similar tests myself. The problem I found was that when the ADC samples the analog signal, the sampling points are frequently at a different position in the waveform than the original digitally copied data from the CD. Therefore, perfect alignment and subtraction is impossible, even if the CD players analog output were a prefect reconstruction of the digital data.

My own conclusion doing this kind of testing is that there is always a margin of error based on the sampling of the ADC. In my case with a sample rate of 96KHz I sometimes found a better subtraction than other times simply by random luck. I'd like to know what methods were used to capture the analog signals in this testing, and how much random variance was found in the first place.

 

I bet they never thought of that.

Despite all the multiple pages of waffle over there they seem to have missed out one mildly important thing- where are the tests to assess at what level they measurable changes become audible?

Without that it's mildly interesting but pointless research.

 

Indeed.

Direct comparison of sampled test signal to source digital data requires perfect reconstruction and then time-alignment of both.

What they do with their funded-and-defence-contracted-'research' is to inject aliases and images into the data, even before the comparison starts.

Small wonder the differences are huge.