Recording at 48 or 44

[Revelation]

Well just a question for you chaps.

Have you noticed any audible difference recording at 48 over 44?

Second question, if you dither down to 44 is it really worth recording at 48?

[cl516]

I second this question.
I've asked before and have gotten a flurry of on and off answers.

[ISedlacek]

Originally posted by Revelation
Well just a question for you chaps.

Have you noticed any audible difference recording at 48 over 44?

Second question, if you dither down to 44 is it really worth recording at 48?

Some time ago I was very much perplexed by these questions and did very thorough practical experiments.When I compared 44 and 48 recordings of identical sources, I have not noticed any reasonable difference.
But when I compared 44k with 96 recording, there was such a HUGE difference (detail, depth, transparency), that never more I will be able to record in 44k (I keep only 96k).
I know , there are many theoreticians around , who will bring 100 + 1 reasons explaining why 96 k SHOULD NOT sound better. Well, for me there is actually only one good argument - to try ... Without that, all such theoretical discussions are pretty useless I think.
Somebody also said that just mediocre convertors sound better in 96 than in 44 just to hide their poor design. Well, I got the same results both with Mytek and Lavry. Are there many reasonably better convertors in the market ?

If a high quality SRC is used, the sonic beauty of 96 k can be very well preserved even after convertion to 44k. I tried most of available software SRCs (ProTools, Cubase ...), the really good ones are only BarbaBatch and Voxengo r8 Brain (freeware !!), without any quality competition.

That 96- 44 conversion is the same seamless as 88 - 44 has been many times proved even theoretically. BTW 96 k sounded better than 88k during my practical experiments

[Tim L]

My setup is:

Initial tracking is digital @ 24/48k -> mix analog -> back to digital @ 24/44.1k. Unless it's going out for mastering (@ 24/44.1), going to 16bit is the last thing I do right before I cook a disc.

Because I go back to analog for mixing, I track @ 48k to capture as much upper harmonics as I can get. In my little world I definitely hear a difference. It's subtle but after going through a conversion to analog for mixing (then back again) I feel it holds up a bit better in the end product.

[LTA]

Originally posted by ISedlacek
I know , there are many theoreticians around , who will bring 100 + 1 reasons explaining why 96 k SHOULD NOT sound better.

I thought all the theoreticians finally came to the conclusion that it was crappy low pass filters in the signal chain kicking in well into the audible sound spectrum that could be bypassed in 96k that were causing the problems.

I personally haven't noticed any signifigant difference between 44.1 and 48KHz either. Some coverters do a better job in that realm than others, of course. With my 3 year old a/d, there is a signifigant improvement going from 48 to 96KHz. Haven't a/bed my archaic a/d at 96K with one of these slick new ones that get past the filter limitations even when operating at 48KHz.

Doing that comparison likely would just give me the desire to upgrade a fairly expensive piece of gear that works just fine already.

[gear chick]

I don't think I hear a significant difference between 44.1 and 48, but I record in 48 anyway.

[NathanEldred]

It's not worth it from a sonic standpoint, especially with good conversion. On a Lavry for instance, there isn't a huge difference between 44.1 and 96k. There is a difference, but it's debateable which is "better".

[gear chick]

Well it's definitly not worth the hard drive space taken up by 96khz even if there is a sliver of improvement, which I doubt.

[SMA]

If you do any sort of digital processing a lot of processors will produce distortion at lower sample rates. This is audible, especially with lots of tracks. At 96hKz these processes tend to be quite a bit cleaner. With clean converters and no digital processing, I'm not sure the sample rate matters all that much in the real world. That said, for classical or clean acoustic I use 96kHz. Hard drive space is relatively cheap and I've got a ton of it.

Oh, you don't actually dither down from one sample rate to lower sample rate, that is "just" a straight sample rate conversion. You dither (or at least should dither) from higher resolution to lower resolution, e.g. 24 bits to 16 bits.

[Wolfgang]

Hard Drives are cheap ... and it is worth it!!!

But as usual, crap in/crap out. Record a nicely miked acoustic guitar in 96k ..... you'll never want 44.1 again.

[ISedlacek]

Originally posted by NathanEldred
It's not worth it from a sonic standpoint, especially with good conversion. On a Lavry for instance, there isn't a huge difference between 44.1 and 96k. There is a difference, but it's debateable which is "better".

I cannot help, there IS a reasonable sonic difference between recording the same source in 44 or 96 even on Lavry . 96 k recording, when converted to 44 k (with Voxengo SRC) and put next to the 44 k reocording has again a bit more air, space and detail.

[gear chick]

I've heard that if I were to record at 96khz that recording at 88.1 would be better because it would dither down better to 44.1 because it's equally divisible by 2.

[Revelation]

With everyone having DVD players to watch movies, and DVD players coming out in cars, it will be very nice to say goodbye to CD's and hello to DVD at 24 bit 96.

[Midlandmorgan]

Originally posted by Revelation
With everyone having DVD players to watch movies, and DVD players coming out in cars, it will be very nice to say goodbye to CD's and hello to DVD at 24 bit 96.

I doubt that will happen within 5 years... and by then who knows what the next big thing will be? Remember DATs? They were supposed to put everything else out to pasture.

[ISedlacek]

Originally posted by gear chick
I've heard that if I were to record at 96khz that recording at 88.1 would be better because it would dither down better to 44.1 because it's equally divisible by 2.

It has been explained many times by experts like Daniel Weiss and others, that there is absolutely no difference between 96-44 and 88-44 conversion. And everybody can practically check it by one´s own ears.

[jho]

I pretty much always do either 44 or 88, depending on what we are going for / project / artist etc. Doing it at 88 seems like less math to me when converting down than if at 96 to 44.

I can most definetely hear the difference in 88 over 44 and 48, and it's clearer to me especially on synths and cymbals.

[Ruphus]

My ears say the same what Ivo said.
It is a difference and a not too subtle one.

96k all the way.

Ruphus

[tonymite]

don't believe ... don't believe the hype !!!

http://www.lavryengineering.com/docu...ing_Theory.pdf

[gear chick]

It's kinda funny. I was reading a thread just like this in another forum and the common opinion was: Just record in 48khz there is no difference and recording in 96khz is actually a detriment to fidelity when dithered down to 44.1.

[Ruphus]

Originally posted by tonymite
don't believe ... don't believe the hype !!!

http://www.lavryengineering.com/docu...ing_Theory.pdf

What hype? We are talking about 96k, not 192 kHz.
Mr. Dan Lavry says around 60 Hz would be optimal. That´s above 44.1 isn´t it?

And I rely on what I hear when I use my convertor at 96k.

Ruphus

[Hiwatt]

You do mean 60khz right? Anyways I think Lavry said around 80khz would be ideal or at least somewhere between 60khz and 90khz. I read this in one very long thread on a yahoo group discussion page.

[Ruphus]

Yes, thanks, there is a little "k" missing.
K like kilo, Kilimanjaro, killer or kilt.
Ladies and gentlemen ... here for you tonight, the K!!!!



Ruphus

[SMA]

Originally posted by gear chick
...recording in 96khz is actually a detriment to fidelity when dithered down to 44.1.

That is simply not true. Assuming decent sample rate conversion, it is going to do a lot less harm than many digital processes that people use on a daily basis. Backing up a bit, though - what do your own ears tell you?

[gear chick]

Backing up a bit, though - what do your own ears tell you?

I don't think I hear a significant change between 48khz and 96, but I haven't conducted extensive listening tests. I think people that say they hear a huge difference should really do this with a blind test. I dont want to repeat what I'm sure has already been stated and A/D conversion 101 stuff, but we can only hear up to 20khz. I know there is all the psycho acoustic stuff, but it's just not a slap in the face difference to me.

[tonymite]

the hype is higher sampling rate = better ... and decimation aint degrading ... u all should learn to read ... i was only emphasising the fact that ... people get sucked in by the 192k thing ... THE DETREMENTAL PART IS THE PART WHERE U PLACE 4 TIMES MORE LOAD ON YOUR CPU TO PROCESS 4 TIMES THE DATA AND THIS MEANS MORE ROOM FOR ERROR ... DONT GET OVERSAMPLING IN THE INPUT/OUTPUT BUFFERS CONFUSED WITH SAMPLE RATES PEOPLE ... AS U MAY ALREADY KNOW ... LAVRY CONVERTERS WIN MOST BLINDFOLD TESTS ....AND ARE THE CREAM OF THE CROP ... SO HE AINT JUST SOME ****** ... HE IS BACKING HIS SHIT WITH HIS WALLET ...

[ISedlacek]

Here we have it again ...
WHat about if instead of "non-believing", and empty theoretical reasoning, you just simply record few identical sources (instruments and vocals) in 44 kHz , then in 96 kHz, then convert those 96 kHz recordings to 44 kHz (through BarbaBatch or Voxengo r8 brain - not through inbuilt SRCs in various DAWs) and simply compare. (it would be maybe useful to have some usable AD convertor for that - not Digi 002 , Delta card etc. ...).
Then share the results whatever they are.
It sometimes happens that those who strongly "don´t believe" in 96 k (etc.) are those who do not have this (or other) facility at home
But anyway - let us rest relaxed and record in whichever way we like.
If we feel like sharing some opinions, it would be nice and valuable if they are based on our direct experience. If somebody really TRIES the above process and feels 96 k sounds the same or worse than 44 k , it would be also a valuable sharing ...
But if somebody only THINKS and speculates, it has not much practical value.

[chrisjin]

in response to the above...

it depends on you A/D...If you use a digi 192, yes you will hear a big difference from 44.1k to 96k...You won't hear as much of a difference if any with the Lavry or the the new apogee AD16X...so what does that tell us?...Its not all about the sample rate.

[ISedlacek]

Originally posted by chrisjin
in response to the above...

You won't hear as much of a difference if any with the Lavry

I tried just few days ago with my Lavry - I was recording one huge string instrument and based on opinons as above, I did recordings in both 96 and 44. Again - 96 k recording had a lot more "air" and overall detail.
It would be a nice relief if 96 k sounds the same as 44, because recording in 96 k brings lot of troubles. Unfortunately it is not so, and as for me - I am ready to sacrifice many things to the sonic beauty. But there is no wonder 96 k sounds better, since it is a double dense information resolution anyway ...

[Boby Stoica]

Here is a point of view on the initial topic, wich was 44k versus 48k.

So, I had a few projects using 48k - and the best sounding SRC I've found was a simple conversion 48k-DA - 44k-AD with my Apogee PSX100, passing the mix through some good analog, wich I always do, but with sessions on 44k all the way.

Recording in 48k - makes a litlle bit of a difference. getting it at 44k doesn't retain much of a bennefit. For me, it's a hassle, I will not do it again and I will ask my clients to bring in 44k sessions if it's possible.

Software synths do sound a bit better in 48k. But converting samples from 44 to 48 it's bad, so , it's a hassle.

[tonymite]

i think u are all ****ed

my " oppinion"