Importance of 96k vs 48k

  • Thread starter Thread starter notCardio
  • Start date Start date
I agree with you Ed... but I find the difference in sampling rates is less striking than the difference in word size...

However, I just did another test last night...

24/44.1 and 24/96 dithered to 16/44.1 redbook -- for the first time I noticed exaggerrated sibilance when I dithered-down the 24/44.1 version compared to dithering down the 24/96 version.... it's not extreme, but it IS noticeable.... the 24/96 dithered version had more transient resolution than the dithered 24/44.1.

This means I may have to re-think my tendency to use 24/44.1 and consider 24/88.2 instead -- I also beleive that the 88.2 makes more sense than 96 - less round-off error when dumbing it down to redbook!

Bruce
 
You know how reputable Hi-Fi magazines review hardware? They set price categories. This is smart because it does not make sence to compare cheap equipment with high end.

Same thing here. In theory higher sampling rate should give more truthful (I am not saying "better") sound reproduction.

In real life it comes down to what equipment is being used. Like Chessrock said, 24/48 Apogee converters may sound better than some cheapo 24/96. In fact the difference should be more vivid at the lower end, than at the high end.

I record at home using computer that I put together myself. I am using Audiophile 2496 and Seasound Solo, and I DO HEAR the difference between 44.1 and 96. My wife hears the difference, my daughter hears the difference. They may have trouble explaining, what is different, but they hear it in blind test. They could tell 100% accurate when I switched recordings of the same song.

Could it be that 48 Apogee sounds better than my best 96? Could it be that in high end studio the difference is practically inaudible? Absolutely! But lets compare apples with apples.

Just 2c from the bottom end.
 
Last edited:
webstop said:
You know how reputable Hi-Fi magazines review hardware? They set price categories. This is smart because it does not make sence to compare cheap equipment with high end.

. . . In real life it comes down to what equipment is being used. Like Chessrock said, 24/48 Apogee converters may sound better than some cheapo 24/96. In fact the difference should be more vivid at the lower end, than at the high end.

. . . Just 2c from the bottom end.
Click to expand...
-----------

I hate to nitpick, but actually the difference should be more pronounced in the higher frequencies, because that is where our hearing is more sensitive (2K-4K).

Whether or not 96K is relevant would probably depend a lot on the content. Consider how important the extreme high frequencies are to the particular piece in question, and how much of a role they play in it's clarity and definition.

If the material is "the piccolo and flute fleet meets the cymbal and bell brigade," then I'm going to make sure my resolution is at the highest possible sample rate - just to be on the safe side.

It may not be as much of an issue when recording "the tuba trio meet the three baritones."
 
You are right about frequencies.
Actualy I meant "low end" versus "high end" equipment, not frequencies.
 
I'm actually in the middle of producing 2 albums right now. At the beginning the engineers and I were deciding on bit and sample rates. Both engineers wanted to do it at 24/44.1 because they felt the dithering process really had to be spot on to get it down to red book, and dithering from 96 is risky. So far everything sounds great, but then again, I'm in a really good studio downtown, so the quality of the gear is better that what you'd have at home. So maybe it's just the killer converters I'm hearing.
 
Well if you are going to record at 24/44.1, you will be doing a lot of dithering. Dithering has to do with bit word size. SAMPLE RATE CONVERSION has to do with converting sampling rates. You have to dither usually after changing sampling rates too!

I would think that if this "downtown" studio could afford the great expensive converters that they could also afford a really nice SRC (sample rate converter) box too! A good SRC box will exihibit NO artifacts converting sample rates.

Ha! Reminds me of a joke.

What do you call an old washed up engineer with bad hearing?














































The "producer".

Ed
 
Then which do you think is better? Record at the higher sampling rate and do the conversions, or just record it at 44.1?
 
"Better" is relative to a lot of different expectations.

If the best audio quality is sought, and the converters are decent, higher sampling rates make sense.

If you think there is a chance that at some point the music will be released on dvd at higher sampling rates, higher sampling rates make sense.

If you are hurting on hard drive space, and/or the cpu is being taxed with the sheer number of tracks, a lower sampling rate makes sense.

I recently did a mix at 88.2Khz 24 bit. The resulting CD from that contained NO artifacts I could hear after the standard SRC in Wavelab was applied. I DID note that the processing I did at "mastering" seemed a lot smoother, which makes sense with higher sampling rates.

Many critical ears agree that higher sampling rates make sense IF you can apply a decent SRC to the mix for getting it to Redbook CD standards for audio. You plain start off with a better sound, and any DSP applied will be better. As long as the SRC works pretty well, I feel it is worth the extra storage to do it. If in the future you want to release the material on dvd at higher sampling rates, you will HAVE something at a higher sampling rate to work with. This is includes multi track AND stereo mixes!

So to recap, the quality edge is with higher sample rates (you will find many "experts" and otherwise who do not agree....you will find that these "experts" are not regarded in the industry or mastering as experts! and those who agree haven't "got it" yet....), but the hardware performance (mainly in PC recordings....) is with lower sampling rates.

Ed
 
ok, eddy... which do you think would sound better?

a) 24/44.1

b) 24/48 with the SRC in a finalizer OR

c)24/48 with your SRC.

xoxo

and no, 24/88.2 is not an option.
 
I would probably just stay at 24/44.1 if 48KHz was the only other option. Not worth the time to sample rate convert with 48KHz files....

Ed
 
You must log in or register to reply here.
Back
Top