Is Dr. Doherty saying that the now common advice to not worry about low recording levels is not good?

[Richard Graham]

Also, the standard of music teaching when I was at secondary school was so bad I don't see how it could get any worse.

Classroom teaching was generally a waste of time, (with some shining exceptions!) There's just so much you can do with "everybody".

Perhaps that's why our music teacher didn't even bother trying. He would turn up fifteen minutes late, tell us he would be back in ten minutes, and return half an hour later, just in time for the end of the lesson. If we were lucky we'd get to sing a couple of verses of The Blaydon Races. I kid you not.

[Richie Royale]

My secondary school music 'teacher' spent all her time making sure that she had a school orchestra and just sent all the others to workstations to play on the terrible Casio keyboards we had (usually trying to play Axel F). The C-Lab computer was hidden away for the one pupil she cherished. I learnt very little in those classes and turned my back on music until I took up making it myself under my own steam.

[Tomás Mulcahy]

Thanks zenguitar and Hugh for the insights on camera tech! Great mythbusting stuff, as always.

[Len]

Can't work out why Dr Doherty would say those things; or maybe we are misreading what he is saying. In any event it seems like poor PR - after reading that I feel less inclined to consider DACS products.

[James Perrett]

Can't work out why Dr Doherty would say those things; or maybe we are misreading what he is saying. In any event it seems like poor PR - after reading that I feel less inclined to consider DACS products.

I must admit that I had similar feelings. If he has strange ideas about digital audio basics then what other areas of audio engineering does he also have strange ideas about and do they have any impact on the sound and reliability of his designs?

James

[Exalted Wombat]

Can't work out why Dr Doherty would say those things; or maybe we are misreading what he is saying. In any event it seems like poor PR - after reading that I feel less inclined to consider DACS products.

I must admit that I had similar feelings. If he has strange ideas about digital audio basics then what other areas of audio engineering does he also have strange ideas about and do they have any impact on the sound and reliability of his designs?

Maybe he just set out to write a "thought-provoking" piece, rushed the job and didn't fully think it through.

[ef37a]

Can't work out why Dr Doherty would say those things; or maybe we are misreading what he is saying. In any event it seems like poor PR - after reading that I feel less inclined to consider DACS products.

I must admit that I had similar feelings. If he has strange ideas about digital audio basics then what other areas of audio engineering does he also have strange ideas about and do they have any impact on the sound and reliability of his designs?

James

Indeed and this;
http://www.benchmarkmedia.com/discuss/feedback/newsletter/2011/12/2/0-ohm-headphone-amplifier-sonic-advantages-low-impedance-headphone-amp
was found to be flawed reasoning I believe?

Dave.

[markhodges]

The lenses used in many old cameras were generally superior to modern lenses.

I find this difficult to believe, considering the technological advances that have been made since then. I would have thought modern lenses are more pure and polished and tested, using computerised equipment:

Most of those technological advances are in manufacturing for the mass market. Also, large format cameras are a very different beast to 35mm SLRs and DSLRs.

If you look at things like the old zeiss folding cameras or olympus trips there really isn't any comparision to a modern P&S camera in terms of image quality. Old large format lenses might be superior to modern P&S lenses in the very limited range circumstances where they can be directly compared but even the best of them will come off worse in any comparison with a direct modern equivalent. There have been advances in materials, advances in design due to widespread computing power, and advances in manufacture i.e. precision grinding of aspherical elements & CNC manufacture of lens bodies, that apply across the board and mean the modern lenses are measurably better.

[James Perrett]

Indeed and this;
http://www.benchmarkmedia.com/discuss/feedback/newsletter/2011/12/2/0-ohm-headphone-amplifier-sonic-advantages-low-impedance-headphone-amp
was found to be flawed reasoning I believe?

I'd be interested to know about the flaw. While that particular article is very marketing led and a damping factor of 6000 is maybe somewhat excessive, low output impedance headphone amps are sold by people like Grace Design as well as Benchmark. While the moving mass of a headphone diaphragm is much less than that of a bass speaker, I would have thought that keeping the output impedance low would only help the sound.

I can understand that a low output impedance has the disadvantage of having more of a volume change between headphones of different impedances though.

James.

[Hugh Robjohns]

I'd be interested to know about the flaw.

I think Dave is referring to THIS THREAD which you contributed to at the end.

I would have thought that keeping the output impedance low would only help the sound.

It does, and is normal practice for high-end products, as you say. The issue was that the tests employed in that report could be argued to be misleading in some of their claims.

hugh

[James Perrett]

Thanks for reminding me Hugh - I don't think that I had followed the link to the Benchmark article in the previous thread so didn't immediately associate it.

James.

[ef37a]

I'd be interested to know about the flaw.

I think Dave is referring to <a href="/forum/showflat.php?Cat=&Board=DESIGN&Number=959384&Searchpage=1&Main=959234&Words=benchmark&topic=&Search=true#Post959384" target="_blank">THIS THREAD</a> which you contributed to at the end.

I would have thought that keeping the output impedance low would only help the sound.

It does, and is normal practice for high-end products, as you say. The issue was that the tests employed in that report could be argued to be misleading in some of their claims.

hugh

Thank you Hugh, that is the very one!

The thing that still bothers me tho is. How can a "zero" impedance source have much of a damping effect when the DC resistance of headphones, in series with it, is usually only a percent or so less than the nominal impedance? If it is argued that it IS different because the resistance is distributed throughout the wire then that flies in the face of all our "black box" models of electronic systems.
Dave.

[dmills]

Yea once Re(Zs) becomes smaller then about 10% of the winding resistance it becomes largely a non issue. The useful form of damping factor of the system is given by the ratio of the motional impedance to the fixed series impedance (including the winding resistance and the amp), not simply by the ratio of the winding resistance to the output impedance of the amp.
Now of course Zs can (and does) vary with frequency (Not by much over the audio band in a good SS amp, but still), which is a good enough reason to measure at 20K as well as 1K for that number.

Interestingly the sort of crude headphone output that has a large resistor in series with the phones across a speaker voltage line never sounded that bad to me, and is in effect a poor sort of current source drive, in that the voltage dropped across the resistor is very much greater then that dropped across the cans (Good current sources appear as high impedances).

Lots of horsepucky talked about headphone drive amps by those with expensive boxes to sell, the things are really not a difficult load by and large.

Regards, Dan.

[Tim Gillett]

On the "using all the bits" theme, here's another example.

A few years ago I was the technical manager of an Oral History audio tape digitising archiving project. We had 7500 hours of recordings to digitise. All the tapes were cassettes, often recorded in a very amateur fashion. The maximum dynamic range of the recordings was unlikely to exceed 60db. Typically it was 30db or less. We had 115db (unweighted) A/D converters running at 24 bits.

I had preset the record levels so that the loudest possible signal ( in this case a Type IV Metal cassette formulation tape driven to saturation - and we almost never encountered one of those tapes anyway) would reach about -3dbFS. So clipping was impossible and the noise floor of the quietest cassette would still be way above the noise floor of the A/D converters, to the tune of about 50db worst case!

Part way through the project, one of my staff (curiously the only staffmember of our group with a formal qualification in Audio Engineering!) stated that sometimes the recordings didnt peak any higher than say -30dbFS - which of course was true - and protested that this represented greately lost resolution. I tried to explain why this was not an issue but to no avail.

I also discovered that such wrong headed notions seem to have infected even the professional audio archiving world. There is a view that we should be digitising even the poorest quality legacy recordings, ie: with the lowest dynamic range, at "at least" 24 bits resolution. The idea of using say 16 bit resolution to digitise a noise laden recording is regarded as selling out on quality.

It's only when you dig a little deeper that you discover the reasoning operating here. It seems that that we are "future proofing" the digitised files in case one day there are developed noise reducing algorithms which can magically remove all sorts of noises from the shocking original recordings, leaving you with a recording far cleaner than ever even entered the microphone, let alone was captured faithfully on the tape at the time.

While there is some truth that certain patterned noises can be reduced, so long as the exact pattern can be isolated and cancelled out with great accuracy, random noises are a different proposition altogether. Since by definition random noises have no identifiable pattern, the software has nothing to work with and can do nothing. It's asking the software to magically identify all of the recorded noises which we as humans dislike and separate them from the noises which we would like to keep such as a voice. It's just not possible.

This is really about noise masking, an often poorly understood phenomenon. If a louder sound completely masks a quiet sound, that's usually it. While technically, the quieter sound has been recorded "underneath", unless we can remove the louder sound which masks it, the quieter sound may as well not be there.

So we have to be practical and realistic in our attitudes to audio, including what's necessary and what's possible.

[Exalted Wombat]

Be very careful with "it will never be possible..." statements.

"Any sufficiently advanced technology is indistinguishable from magic."

[Shivanand]

The sound of 'analog' is simply what people have become used to; what they heard first in their formative years will colour their perceptual interpretation later on....

That is very true. However, I'm one of those who grew up on analogue - but I can see and hear that digital sound is an incredible improvement on that awful wow and flutter (I have a keen sense of pitch) and all that mud that anologue gives us.

+1 to that.

However, the musicians are rarely as good these days.

That may apply to many "popular" musicians but, come on, there are many superb musicians around these days.

[MonkeySpank]

Be very careful with "it will never be possible..." statements.

"Any sufficiently advanced technology is indistinguishable from magic."

Correct. CEDAR Audio already produces audio restoration processors that do an astonishing job of de-noising and de-popping old recordings. And if Melodyne can make a plugin that can separate individual notes from a chord, and Photoshop can offer a plugin to de-blur a photo, then everything will possible some day.

[narcoman]

Both done perfectly!!

[Andi]

I've noticed that Eastenders has been crap since we got a digital telly!

[hollowsun]

I've noticed that Eastenders has been crap since we got a digital telly!

I noticed that Eastenders was crap when I had an old CRT. I don't have a TV at all now and Eastenders is still crap!

The only good thing about Eastenders is the Simmons tom fill!

[Tim Gillett]

Be very careful with "it will never be possible..." statements.

"Any sufficiently advanced technology is indistinguishable from magic."

Correct. CEDAR Audio already produces audio restoration processors that do an astonishing job of de-noising and de-popping old recordings. And if Melodyne can make a plugin that can separate individual notes from a chord, and Photoshop can offer a plugin to de-blur a photo, then everything will possible some day.

There are such things as physical limits. The best audio restoration processors do a great job within the limits of what is possible.

Any fool can use Photoshop to deliberately blur an already sharp photo. But you cannot "deblur" a photo because you cannot "uncover" photographic detail that was never captured in the first place.

Example: the 1969 Apollo moonwalk videos. For the 40th anniversary in 2009 NASA wanted to present much clearer vision for the public than the low quality stuff we had always been used to seeing. The only way to greatly improve the vision quality was to find the 1" "slow scan" data tapes which had much more picture information on them than the low quality kinescope stuff re-recorded off TV monitors at the time. Well they couldnt find the data tapes. Maybe one day they will find them. Then and only then will we see significantly better moonwalk video images.

Now, an example where it did work. Why does the "restoration" of the 1939 (I think) " Wizard of Oz" look so (relatively) good compared to earlier jobs? Because the original 3 strip black-and-white Technicolor negatives survived. Where software came in was in accurately aligning the three separate images to one another - registration - restoring them to one sharp and clear colour image. That had not been possible previously. But someone had had the foresight to see that one day it might be possible to accurately align them and so they had carefully preserved the separate negatives all those years.
The difference was that the basic information had been preserved. Without that you can do nothing. But with it, yes great things are at least possible.

Another; Why do some old US TV shows look so good on DVD re releases today? Why do they look actually better than what we remember them looking like back in the 60's when some of us first saw them?

A combination of factors. The shows were originally shot on 35mm film stock which has been kept and preserved. Even the original camera negatives were kept, rather than just copies. The original magnetic audio tapes had also been kept. 35mm film at the time had much greater resolution than broadcast TV at the time. Much better electronic film scanning hardware is available today than back then. Much better consumer release formats, like DVD and Blu ray are available today. Higher resolution TV monitors are available today.

All of those things combined mean potentially great picture and sound today for the consumer. Not some piece of "magic" software on its own. The software always has to have something to work with. The quality has to be there in the first place.

Sadly, there is a view today that we can "restore" elements of an audio or visual recording that were never there in the first place. A sort of "garbage in, perfection out" mentality.

Then there's the equally false assertion that by strengthening the already strongest link in the audio or visual chain, you are somehow strengthening the weakest link in that chain. But the chain is only as strong as its weakest link, whatever that may be.

[hollowsun]

The sound of 'analog' is simply what people have become used to; what they heard first in their formative years will colour their perceptual interpretation later on....

That is very true. However, I'm one of those who grew up on analogue - but I can see and hear that digital sound is an incredible improvement on that awful wow and flutter (I have a keen sense of pitch) and all that mud that anologue gives us.

+1 to that.

+1 to your +1!!

I think tape has a certain charm in this day and age and can be useful as an 'effect'. Back in the day, of course, it's all we had and apart from dewy eyed nostalgia, rose coloured glasses and all that, a £500 digital 'portastudio' will outperform an old MTR that cost many £thousands in 'ye goode olde days'.

However, the musicians are rarely as good these days.

That may apply to many "popular" musicians but, come on, there are many superb musicians around these days.

Absolutely true. There ARE some quite deluded individuals out there, though, has be said.

How's the Mazda, Shiv?!

[Exalted Wombat]

However, the musicians are rarely as good these days.

On average, you're quite right! But EVERYONE records now. Only the good ones used to get anywhere near a microphone.

[Tui]

Example: the 1969 Apollo moonwalk videos. For the 40th anniversary in 2009 NASA wanted to present much clearer vision for the public than the low quality stuff we had always been used to seeing. The only way to greatly improve the vision quality was to find the 1" "slow scan" data tapes which had much more picture information on them than the low quality kinescope stuff re-recorded off TV monitors at the time. Well they couldnt find the data tapes. Maybe one day they will find them. Then and only then will we see significantly better moonwalk video images.

You really believe all that stuff NASA is telling us, do you? Like, the biggest achievement of mankind, ever, was documented and broadcast not directly via a live feed, but re-recorded from images that were projected against a studio screen? And even better, the only recordings that supposedly were of higher quality, somehow and tragically, were misplaced by a hapless NASA employee?

[Tim Gillett]

Example: the 1969 Apollo moonwalk videos. For the 40th anniversary in 2009 NASA wanted to present much clearer vision for the public than the low quality stuff we had always been used to seeing. The only way to greatly improve the vision quality was to find the 1" "slow scan" data tapes which had much more picture information on them than the low quality kinescope stuff re-recorded off TV monitors at the time. Well they couldnt find the data tapes. Maybe one day they will find them. Then and only then will we see significantly better moonwalk video images.

You really believe all that stuff NASA is telling us, do you? Like, the biggest achievement of mankind, ever, was documented and broadcast not directly via a live feed, but re-recorded from images that were projected against a studio screen? And even better, the only recordings that supposedly were of higher quality, somehow and tragically, were misplaced by a hapless NASA employee?

Yes I do believe it. Understand that the original video quality sent from the moon was well below the video broadcast standards of the day, including being only black and white. The Apollo 11 camera sent images of 320 x 250 lines at only 10 frames per second, well below broadcast standards of the time.

So even if the tapes were found and processed using modern digital conversion software, they still wouldnt be great quality, just better than what we saw after the necessary conversion to broadcast standards for worldwide viewing. All they would be doing is removing one weak link in the chain, the primitive conversion at the time, improving the picture quality significantly, but not magically turning them into something they never were.

By comparison the colour still photos taken on Apollo 11 were far superior in image quality. You would expect that. They were taken with a Hasselblad 6 x 6 camera and the original camera films came back to earth.

Re losing the original data tapes, you might be surprised how commonly even top audio, film and video original elements were trashed or recorded over. A significant proportion of the BBC's TV archives from before the late 70's was lost, partly because before home video really took off they saw no further money to be made from reissuing them.