Recording at 24/96

[Macinbowl]

So, with the growing popularity of high definition format, how many of you (cause I didn't yet) have switched to an all HD audio signal chain at 24/96 ?

If not, for those who have to delivered their project in HD format, what are your project "workflow" regarding of bit/KHz?

Recording/mixing: 44.1? 48? 96? 192?
Bouncing/encoding: 44.1? 48? 96? 192?

Thanks!

Mat

[resolectric]

I am recording at 88.2 with 32 bit Floating Point (Nuendo 3).
I keep all further processing in the same Sample Rate and Bit depth, mix the same way and deliver BWAV files to Mastering just like that: 88.2/32.

Sometimes i work at 96K but not very often. I've read too many comments on how downsampling to 44.1 from an original file at 88.2 is safer than from a 96 original file, so, i do it until some better arguments suggest 96KHz is 100% safe when downconverting to 44.1KHz.

Whenever i receive a Mixing job with files that were recorded elsewhere and if they are 44.1 or 48 i upsample them to 88.2.
I find that processing applied to those files sounds fuller (richer, more defined... adjectives for better sounding) when the files have a higher sampling rate. This is far more noticeable with processing that causes a "tail" in sound, such as Reverb or Delay.

So, i never dither and never downconvert.

[Henry Robinett]

Mostly I record/mix at 44.1, but sometimes I record mix at 88.2. The project I'm doing now is 88.2. The only headache I have is my inability to use my beloved PCM91. If I really need to I open an empty project and record the reverb at 44.1 and then convert it.

I have the ability to go 192, but haven't as of yet. I'm about to do some reviews of some microphones and will record them at 192.

When I'm at 88.2 I stay there until either I master it, or someone else does. .

[Phil O]

I've read too many comments on how downsampling to 44.1 from an original file at 88.2 is safer than from a 96 original file, so, i do it until some better arguments suggest 96KHz is 100% safe when downconverting to 44.1KHz.

You may have been given incorrect information. It seems logical that going from 88.2 to 44.1 would be simpler than going from 96 to 44.1, but that logic is flawed. One must consider the algorithms used to do the conversions. It's not a simple division and the fact that 88.2->44.1 looks simple but 96->44.1 doesn't has nothing to do with it. I've not seen any information (that doesn't rely on speculation) that suggests that one is "safer" than the other.
The real question that should be asked is whether or not higher sample rates are worth it. I'm not even going to climb into that can of worms except to say that I work at 44.1K, 24 bit most of the time. I will work at other bit depths/sample rates if the client requires it keeping this in mind - do as few conversions as possible over the course of the project.

Phil

[Henry Robinett]

There is absolutely no question to me that the higher sampling rates sound much better. Beyond all theoretical discussions is "how does it sound TO YOU?"

No question that TO ME 88.2/96 clearly sounds much clearer and well defined than 44.1/48. No question to these ears. Now as to the benefit to something that will be down sampled t 44.1/16 is another question entirely.

[bongo_x]

24/44.1

sometimes 16 bit.

bb

[Macinbowl]

do as few conversions as possible over the course of the project.

This is the key. Recording at 24/96 and work without conversion 'till the end of the project is the way to go.

I do a lot of Live show recording. Some of them are for HD product like HD televison broadcasting or Blu-Ray disc production. Unfortunatly, most middle cost digital live console still work at an internal SR of 48k so the use of digital output card make it imposible to keep an all digital path at 24/96 from the recording session all the way to the final product.

I'm pretty sure that some guys out there are still recording at 24/48 and then upconvert their stuff at 96 after mixing...

[jrdmcdnld]

24/44.1

sometimes 16 bit.

bb

Works every time!

[resolectric]

I've read too many comments on how downsampling to 44.1 from an original file at 88.2 is safer than from a 96 original file, so, i do it until some better arguments suggest 96KHz is 100% safe when downconverting to 44.1KHz.

You may have been given incorrect information. It seems logical that going from 88.2 to 44.1 would be simpler than going from 96 to 44.1, but that logic is flawed. One must consider the algorithms used to do the conversions. It's not a simple division and the fact that 88.2->44.1 looks simple but 96->44.1 doesn't has nothing to do with it. I've not seen any information (that doesn't rely on speculation) that suggests that one is "safer" than the other.
...

Yes, it might happen that i may have read incorrect information but it seems that you are aware of that information as well. You must have read it.
I recommend that you read, as an example, the explanation behind the functioning of the algorithm on the R8BrainPro from Voxengo.
It is recognized as one of the best (if not THE best software SR converter available) and it uses a principle based on the conversion through simple mathematical multiples. It first converts the file's Sample Rate to a value that is then easily downconverted to the required SR.
And it sounds better. For some reason.

Anyway, yes, you may be right. But since i'm not wrong either, we can keep doing things each its own way.

The real question that should be asked is whether or not higher sample rates are worth it. I'm not even going to climb into that can of worms except to say that I work at 44.1K, 24 bit most of the time. I will work at other bit depths/sample rates if the client requires it keeping this in mind - do as few conversions as possible over the course of the project.

Phil

I've never heard anyone that uses 44.1 saying that they hear a difference between 44.1 and 88.2.
On the other hand, everyone i know of that uses 88.2 does it because they hear a difference.

I know i do.
And to my hears and in my studio, 88.2 is clearly better sounding than 44.1. Blind test or not, 100% of the time.

[jrdmcdnld]

Forgive me if someone mentioned this already. I believe you should always sample at the rate of your final product. SD DVD - 48k. CD - 44.1k....I think that no matter how good of a sample rate converter you have, you should avoid that process.

[conleycd]

I used to record at 24/96 but I have been consistently using 24/44.1khz now both to avoid downsampling and to free up DSP resources in order to maintain "live" plug ins. I feel comfortable at this level given that the majority of Digidesign HD users are recording in the 24/44.1khz format. I am also reassured by people like Tom Lord-Alge (mild mixing idol of mine) who has for many many many years recorded and mixed in 16/44.1khz (on Sony digital tape). In fact, during mix projects, he transfers any higher bit higher sample format 'down' to his Sony digital. If he can sound great... maybe I can too (I guess I just need to save up a little more a SSL E-class board like he has).

CC

[jrdmcdnld]

...yep, I've heard that about Tom too. I believe he stated that he doesn't trust hard drives and he doesn't like the way they sound either....as though hard drives have a "sound".

[conleycd]

Yeah... strange - Tom. However, no matter how weird of a thought it is that he thinks hard drives have sounds (or Eric Johnson believes fuses and which way a resistor is install in an amp affect the sounds) there is no doubt that these somewhat strange people with strange beliefs make a lot of money doing what they're doing. So while I might not always agree with them, I do respect them. I think the truth is (with both people I cited) their skill transcends their equipment (even if they themselves don't believe it).

It's the same with Tom Scholz. The Rockman was sooooo compressed it was hard to distinguish tone from anything that came through it but he was nuts about the way things sounded and had all sorts of strange beliefs (now he is a professional engineer) but still....

CC

[monkey man]

24/44.1
As you'd expect, the rate and bit depth are maintained through all project phases.

My sources are modelled electric (and one day acoustic too) guitar through a POD and DI, basses through a POD and DI and vocals recorded from a dry booth. I don't see any point whatsoever in upping the sample rate for these sources. If I were using a good mic in a great room to record almost any acoustic instrument, I suppose I'd reassess the situation because the integrity of all those real harmonics and room behaviour could conceivably benefit from higher precision at the capture stage.

[kassonica]

Yeah... strange - Tom. However, no matter how weird of a thought it is that he thinks hard drives have sounds (or Eric Johnson believes fuses and which way a resistor is install in an amp affect the sounds) there is no doubt that these somewhat strange people with strange beliefs make a lot of money doing what they're doing. So while I might not always agree with them, I do respect them. I think the truth is (with both people I cited) their skill transcends their equipment (even if they themselves don't believe it).

It's the same with Tom Scholz. The Rockman was sooooo compressed it was hard to distinguish tone from anything that came through it but he was nuts about the way things sounded and had all sorts of strange beliefs (now he is a professional engineer) but still....

CC

Yea Eric Johnson had proved to many people that he can hear difference between evereadys and duracell 9volt batteries in his pedals. His choice is Duracells.

I do believe some people can have this ability.