96k - 002

[wez]

just the two will do me :-)

[mfdu]

hey wez-o

so as far as i gather the s/pdif badged digital i/o will handle 24/96 - just make sure you set the receiving units clock to slave.

two channels out of the 002 at 24/96 shouldn't be a problem. it's the multitracks that take a bit of work.

other option is to "export selected regions as whole files" if you just want the raw audio . . .

what do manny's say?

:)

chris.

[rick]

i am still waiting for somebody somewhere in the world to tell me that they have sent 96k digitally out of a 002
the one here will not do it , with any converters or any combination of tricks and tales
my vote is it appears not to !

[mfdu]

rick

if that is the case, it would indicate that i wont be able to use the s/pdif with an extra two channels of ad/da to increase my input count.

scary. but i can't help you from a "yes i have done" perspective, i'm sorry.

surely someone can pitch in and resolve this?

wez - what did the manny's crew say?

chris.

[ottowr]

I think there's more benefit in using 24 bit than using a higher sample rate.

You can use/burn 48k or 96k 24bit LPCM on DVD-Video discs.
I have a James Taylor "Live at the Beacon Theatre" DVD that is 48kHz 24bit LPCM, and it sounds much more detailed than CD or other DVDs.

[Kris]

That is one sweet sounding dvd. I agree.

[rick]

if anybody out actually see this and gives a toss after all this time

the digi 002 DOES output 96k on the spdif (rca) port , but it took me at least 7 different converters and months of on again off again trials and web scanning to make it work ( as well as bugging the hell out of howard jones for no financial reward to him :( sorry howard :)

it seems whatever the standard for spdif WAS
IT IS NOT THAT NOW !

so if you have a NEW 96 k converter that will see fast 1 wire hes res digital 96k & up .
the the digi 002 will lock up and transmit 96k into it no problems
i presume it will record 96k digitally as well but i have not tried that.

it worked for sure into the "benchmark" converters i tried today


i get nearly no 96k stuff from the prostudios with big megabuck hd tools rigs
its all 44.1 or 48 from them
but i get a smattering of 96k sessions from home studio guys who have 002/003 rigs .



if you wondering what sample rate bit rate we actually WANT to see - we can handle anything but given a choice 24 bit 44.1 k wave files is what we would like to recieve



i have half a dozen or so "pro" guys who use to do things at 96k , but have decided that digidesign sounds BETTER for rock n roll at 24 bit 44.1 or 48k

why i can only guess ?

but if you have tried 96 recording for a number of sessions then you go back to 44.1 or 48 against all number crunching nerdy logic it must have something to do with the way it sounds

i have had a bizzare amount of dsd ( sacd ) interest in the last few weeks
( this is something we do not currently provide)
so i was looking into buying some new converters to make dsd happen
and i figured while i was testing things i would solve this "does 002 output 96k digitally or not " issue .

i have looked at over $40 k worth of converters in the last week .

err... i can wait a bit longer to get into dsd :)
i thought it had gone away without costing me a cent but apparently its THE thing in japan this year.

i had thought blueray was the thing in japan ....bugger ... more converters !

[Howard Jones]

If the intended release format is CD then why oh why would anyone record at other than 44.1 or 88.2kHz? Getting from 96 to 44.1kHz must do wonders for the sound of the tracks...

[Mark Bassett]

Why? Well the anti-aliasing filters found on the inputs of cheap A/D's can create artifacts going down to about 11kHz when recording at 44.1k and can therefore sound much worse than using sample rate conversion with a nice smooth roll-off filter.

The end result is that a low cost 96k A/D through a decent SRC can produce better sounding end results than a low cost 44.1k recording with no SRC.

[Howard Jones]

Yeess, but if someone can't afford a decent AD to start with, how can they afford to be involved with 96k gear?

When I asked the question, I was thinking more like it sounded that the clients Rick was referring to are using 96k by choice without having thought through the consequences.

[Howard Jones]

And a further thought... surely a file recorded at 88.2kHz with a relatively simple down-conversion to 44.1 is going to sound better than a file recorded at 96kHz and down-converted to 44.1?

And yes, I am an official grumpy old bugger who is happy to assert I'll bet there isn't a soul alive who could pick the difference between the native sound of a file recorded at 96kHz and one done at 88.2kHz.

[Ausrock]

And the end product is likely to be squished down to the point of being an anorexic shadow of it's original self/source just so it can be crammed alongside a thousand other songs into something not much bigger than a credit card.

[Chris H]

And the end product is likely to be squished down to the point of being an anorexic shadow of it's original self/source just so it can be crammed alongside a thousand other songs into something not much bigger than a credit card.

And this points to the playback device being more important. in marketing tearms, than the content..........apple is into selling iPods with the synergy of marketing music downloads, as Macdonalds is into realestate via selling crap food hamburgers.

[Sammas]

If the intended release format is CD then why oh why would anyone record at other than 44.1 or 88.2kHz? Getting from 96 to 44.1kHz must do wonders for the sound of the tracks...

I've been lead to believe that it depends on the type of sample rate conversion used. It isn't always simply a case of straight down sampling (ie. halving 88.2khz to get 44.1khz). Some sample rate conversion up-samples to the lowest common denominator of 44.1khz, 88.2khz, 96khz extra then downsamples to the chosen final sample rate. There are two main types of downsampling - downsampling by integer factor(ie, straight down) and downsampling by rational fraction (ie, sample up then back down).

On a 002, yes there is a big difference between the sound at 44.1khz and 96khz. On a digi192 or apogee rosetta800 I can't notice any difference. I assume its due to the design of the sinc filter in the converters. The sinc filter has a significantly easier job at higher sample rates than lower ones, but a properly designed and implemented since filter (ie. in apogee gear) will work very well regardless of the chosen sample rate. In lower end gear however, it probably becomes a whole lot more noticable as the sinc filter suffers from the increased workload at 48 & 44.1khz.

[mfdu]

i think you may be on the money with that one. because i keep swearing blind that the reverb tails are too fractured on the 002 at 44.1, but a lot smoother at 96.

feel free to point out that it's because i've got a 002. fine. i don't care.

chris
mfdu

(sulking in corner with arms crossed)

[Howard Jones]

Fine. But I'm still saying that 88.2 will also be good when compared to 44.1, like you are saying that 96 sounds better than 44.1. I'm a bit intrigued that 88.2 has become the forgotten sample rate in favour of 96k. Can 002 actually run 88.2?

I take the points about filter techniques and methods of decimation. It's interesting because, in early days, the filters were one of the single greatest weak spots and multiplication followed by decimation was seen as a great evil. Have things really improved that much? Or has the marketing gotten in the way somewhere here?

It's curious from another point as well: 48 exists solely because it was developed for video applications. Doubling the sample rate to improve things lead to 88.2 for audio and 96 for ... what? 96 has become the Fs of choice for audio and video applications don't touch 96k. All very interesting.

[Kris]

Very true Howard.

I'm currently editing on a state of the art system. 10 bit uncompressed HD video, 24 bit audio at 48khz. That's as high as it goes. Even HD video recorders (the $150,000+ ones) don't go past 48khz. So if you're doing a blu ray dvd....guess what? You're gonna be upsampling.

[mfdu]

i've never used 88.2 on the grounds that - if i'm going to pay the price for higher sample rates (in terms of storage and archiving) then why be half-assed about it? go on - go the full monty.

but i'm finally hearing you on the downsampling issue - 88.2 / 2 = 44.1, clearly. i've been purposefully focussing on the "but it sounds better than 44.1".

i'll do some tests of 44.1 generated from 96 vs. 88.2, in my own flawed environment, with my own flawed perceptions, and i'll see how wrong i have been all along.

but i'm not sulking!!!!!

:)
chris
mfdu

[Howard Jones]

No, no - you may well be right. It's just that, as a paid-up sceptic, I always sniff suspiciously around activities that appear to have become received wisdom kinda out of nowhere. 96k just largely arrived as a done deal and the tone of discussion on this topic in this thread has confirmed my suspicions that it may have had a marketing component involved somewhere along the line i.e. at some stage in the recent past everyone just went over to 96k automatically and we never saw any detailed arguments about 96k vs 88.2k.

It has me wondering how much gear will actually do 88.2 as well as 96k. Have the manufacturers abandoned it? Looks like the marketplace has.

[wez]

pretty much anything that does 96 will do 88.2

and at the risk of revisiting a topic that i have read/argued/lost sleep over to the point of brain implosion.... it should make no difference going from 88.2 ->44.1 or 96 ->44.1. if there is it means that the SRC is broken. the apparent simplicity of the maths is entirely irrelevant.

if you really want to do your head in go on over to Glenn Meadows' mastering web board where this topic has been done to death many times over by people who actually know about this stuff (including the dan lavrys & michal jurewiczs... but that's a whole other thing to kill yourself slowly by). after that you'll never want to spend another moment thinking about it again. ever.

[Howard Jones]

So, people use 96k because it sounds better? Or - more likely - they use it just because they can.

[Ausrock]

Or like me...........I don't use it :-)..............I'm more than happy to stick with my D160's 16/44.1 or 1" R2R.


And Howard you may be right about 88.2 being passed over as it's only quite recently that I'd even heard of it, although that may be partly my fault as I've made a point where digital is concerned of only learning what I needed to to suit my needs, anything else has been regarded as surplus drivel and consequently ignored.

:-)

[daniel_c]

So, people use 96k because it sounds better? Or - more likely - they use it just because they can.

If I am not mistaken, 96kHz is part of the DVD spec that was really marketed heavily and hyped in many advertisments. That's probably where the bias toward 96kHz is coming from. 88.2kHz is just as good. Now the DVD spec allows for various sample rates, but I remember 96kHz being touted as the be all and end all for DVD audio.

I hate linking to Wikipedia, but here I go; http://en.wikipedia.org/wiki/DVD-Audio

[Howard Jones]

Interesting point. It's all ancient history now but, prior to the inception of DVD, there was quite a dogfight over whether DVD should feature multi-channel sound or high-res sound, as there wasn't sufficient storage space to cater for both.

In the crudest of terms, there were 2 camps: the American that wanted multi-channel and the Europeans who wanted high-res. The US campaigners were led by Tomlinson Holman of THX renown and one of the leading Europeans was Bob..Bob... forgotten his surname, from Meridian. The latter went on to develop MLP (Meridian Lossless Packing), I think as an answer to the American complaints that high-res sound would chew up too much disc space.

I see MLP occasionally here and there, but I don't know that many products have implemented it.

The Americans saw audio on DVD as an adjunct to the video and that therefore multi-channel sound was what was wanted i.e. they wanted DVD to replicate the theatre experience. The Europeans saw DVD as an opportunity to take audio to the next level. Sony muddied the waters by introducing SACD which I thought was pretty much dead. However, discussions with Rick this week have caused me to reconsider.

Coming up we have this decade's shitfight between HD-DVD and Blu-Ray. Great.

[rick]

i believe in due course we are all going to find out that the quality of the highest end converters
ie the prisms, the apogees , dad , lavrys weiss, and all the rest are all infact flawed designs in the analog part of their circuit design, and infact a $10,000 converter is really no better for your music then a $10 converter.

but you going to have to wait untill somebody figures out ( as they have in the boutique mic preamp world )
that people will pay for " true sound improvements" and that may never happen of course
right now people pay for "marketed" sound improvements and changing standards and formats render each 5 year old product "ancient history "

right now people pay for the highest numbers they can get , qualified by price , qualified by an impossible position where they can never actually be checked out in a real a/b/x test , qualified by they cannot actually work their system with out the new interface on the new converter (ie firewire etc etc )

if you think you have trully compared converters , i think you are wrong

i have tried bought , listen to and gotten pissed off at more converters and more music through converters then anybody i can find to talk to on the subject
sure the number crunching in the programs has gotten better ( and it needed to ) and maybe it all works better at the higher numbers maybe not , but i am sure its not that simple , sample rates are but one of the gazillion parameters going on

i would go so far as to say any improvement you have heard from any converter that sounded better then 44.1 has nothing to do with the sample conv. rate,

it has to do with the quality of the real world analog circuits attached to it , the power supply that circuit is given and the output drive circuit of that circuit and the way the whole system locks together
( lets not talk about work clocks today i could not bear it )


for a brief time i thought converters where going to be a cool thing , the david manley company ( as opposed to the current manley company built a converter with real analog outs and it was and is the only converter that i have ever experienced that made me believe in converter "magic" it was a real audio box not some game boy pretender like the ones we all use daily

while i was saving up to buy the manley gold ref converter ( its was $20,000) the "bit rate wars started" and his converter ( which was a marketing flop) was discontinued and never revisited.
if anybody finds one on evil bay please tell me first

[JulienG]

It isn't always simply a case of straight down sampling (ie. halving 88.2khz to get 44.1khz). Some sample rate conversion up-samples to the lowest common denominator of 44.1khz, 88.2khz, 96khz extra then downsamples to the chosen final sample rate.

It shouldn't ever be. It is a little non-obvious, but you can't simply drop every other sample and get a perfect 44.1 conversion. To some extent it's similar to the whole dithering thing with bit-depth.

One way to think about it is with what happens to a square or triangle wave with an attrition SRC you'll see that you need something better really quickly.