Audio measurements matter!
- Thread starter Fatmarley
- Start date
Dougie2404
cranky old git
Has any of us listened to a measurement? If I decide to just what part of the music should I listen to and for what.
Purité Audio
Trade: Purite Audio
Erin’s audio corner, excellent reviews, Erin purchased a hugely expensive Klippel NFS measurement rig.
Keith
Keith
gustav_errata
pfm Member
I haven't watched yet, but does anyone actually say measurements don't matter? If anything I think the most common refrain would be that measurements don't tell you everything, such as how you will perceive the output of a device.
I don't think anyone would argue that measurements don't matter during design of the device.
I don't think anyone would argue that measurements don't matter during design of the device.
Rug Doc
pfm Member
Measurements do matter.. I tried to EQ my system (Using the EQ available in Roon) by ear and got it sounding good to my ears.
Last week I plugged a UMIK mic into REW and did a sweep, then used a Harmon house curve and used the REW EQ settings to give some EQ suggestions.... Plugged these into Roon and 'Bingo' the system has never sounded smoother, the stage has got deeper with easier front to back definition and the sound is just more realistic, so it's worked for me..
Audiolab 8200CDQ DAC, Audiolab 8200P Amp, Impulse horns.
In the car audio world, measurements and tuning is imperative... equipment is 10% of the final result - tuning is 90% it makes that much difference in this situation as the drivers are so far apart.. mids in doors, tweeters on dash and sub behind you in the boot - you NEED time alignment and EQ to sort it all out.. Try to do this without an RTA or REW and you'll struggle..
Last week I plugged a UMIK mic into REW and did a sweep, then used a Harmon house curve and used the REW EQ settings to give some EQ suggestions.... Plugged these into Roon and 'Bingo' the system has never sounded smoother, the stage has got deeper with easier front to back definition and the sound is just more realistic, so it's worked for me..
Audiolab 8200CDQ DAC, Audiolab 8200P Amp, Impulse horns.
In the car audio world, measurements and tuning is imperative... equipment is 10% of the final result - tuning is 90% it makes that much difference in this situation as the drivers are so far apart.. mids in doors, tweeters on dash and sub behind you in the boot - you NEED time alignment and EQ to sort it all out.. Try to do this without an RTA or REW and you'll struggle..
Robert
Tapehead
Take for example, the MM phono input still found on many amplifiers.
There is no standard for cartridge loading other than it should be 47K Ohms and manufacturers frequently deviate from this - and capacitance is a total crapshoot.
Then we have the accuracy of the RIAA correction, which can vary by several dB depending on the design.
All can have a very audible effect on performance.
You (or rather, technical reviewers) can measure and understand this before purchase, or you take pot luck and throw money around hoping to get lucky and not really grasping what's going on.
narabdela
who?
Sue Pertwee-Tyr
Accuphase all the way down
Me, too. I think the initial premise is false, as gustav points out.
I remember once speaking to a technical reviewer. He reckoned he could give a pretty good steer on how something would sound from the technical measurements alone, and unsurprisingly, his subjective reviews invariably came out that way. It seems to me there's the potential for bias in that process. Also, the process becomes self-confirming. You measure this, and you correlate it to that sound signature and because you always hear that when you measure this, there's no reason to challenge the association or question whether the link is causation or simple chance correlation. Measurements are an engineering tool and are essential in that context, but used elsewhere there's a risk that you're using a spanner as a hammer - just the wrong tool for the job. Using them to inform a subjective review is fraught with similar risks to those that sighted, subjective reviews are criticised for, namely the possibility that bias creeps in.
Tony L
Administrator
Agreed. The thing that winds me up more than anything is the patronising position that because someone exists maybe more to the subjectivist end of the spectrum they don’t understand the measurements etc. Many of us have likely forgotten more about audio history and technology than some of this new bunch of internet crusaders will ever know!
Robert
Tapehead
Agreed, and it's important to understand his audience, and ours.
We sit in a niche of a niche, mainly a lot of older audiophiles primarily interested in simple 2 channel audio. We've been through a near lifetime of experience in a period where measurements were largely pushed aside, from the late 70s onward the subjective reviewer was king.
You only have to read the two threads currently running for the best and worst amplifiers. The same products crop up in both threads with completely disparate opinions, because there are no reference points and stories propagated through the years about this or that brand colour the opinions.
Erin has a younger audience with far more interest in HT and no fear of all things digital, including DSP.
He isn't discounting subjective opinion - that would be madness since you like the sound of something or you don't
- but he is right to drivr home the message that measurements matter.
Sue Pertwee-Tyr
Accuphase all the way down
Measurements are essential, as an engineering tool during the design and development process; and for repairs, etc. It doesn't follow that they are necessary as part of the subjective evaluation process. It's like somebody says 'I've got this box of tools, so I'm going to use them', without recognising that they may not be the most appropriate thing for the circumstances. As somebody once said 'if all you have is a hammer, everything looks like a nail'.
Arkless Electronics
Trade: Amp design and repairs.
I was going to post on the lines of only a complete moron would consider measurements unimportant.... your post kinda covers it though.
It's not possible to design hi fi gear without measurements!!
For the purchaser of hi fi they are still important in as much as it allows you to avoid the real stinkers.
As I keep saying on this subject, an item that measures as having a roll off below 70Hz or above 9KHz or has 6% distortion is going to sound very obviously wrong. Once we are at say 10Hz - 30KHz +/- 1dB and less than 0.1% THD then it measures well and the ears must take over as the final arbiters.
Sue Pertwee-Tyr
Accuphase all the way down
Fair point, but how many products actually make it to market with numbers like that? It does seem to me that most products measure ‘acceptably’ and then we’re into discussion of what degree of ‘better than acceptable’ is required. I do sometimes feel that the measurement contingent will, on the one hand, argue that a measured criterion which is 0.05% is ‘better’ than one at 0.1%, and will affect the sound ‘like this’ while simultaneously arguing elsewhere that a difference of 0.05% is likely to be inaudible.
Arkless Electronics
Trade: Amp design and repairs.
Thankfully very few do but there are a few SET's for example that measure that badly and they will be more "sound effect generators" than true hi fi.
Probably of more relevance would be "at one point do measurements guarantee that the subjective performance is free from correctable errors which could be measured and assessed". Even 0.1% THD could be too much if it is largely odd order distortion at high harmonics. Human hearing range is usually accepted as roughly 20Hz to 20KHz (for a youngster anyway!) but if we brick-wall it at this it will be noticeable to some degree. I make all my amps flat down to <5Hz for subjective reasons for example.
Dougie2404
cranky old git
I agree that they matter in the design but the result can still sound crap. Surely it's the end result that matters.