Of bits and other things

[TBman]

I had found tonight that I had changed the input recording to 96000hz and the bits to 32 in Audacity. Turns out this is not a good idea on my system as 44100 and 24 bits sound better. I was going through a test of playback of a mp3 of Al Petteway's playing and recording it, to see what's up with my recording gear when I found this. I also noticed that my recording gear (two condensors, a mxl 991 and AKG perception 150 as a spaced pair into a usb tascam us-122) introduces a bit more bass than what appears in the original mp3. Any idea what could be causing this? Cheap mics, cheap Tascam? The room?

[rick-slo]

? Restate the question.

[runamuck]

Originally Posted by rick-slo

? Restate the question.

Yes. Ask again.

[TBman]

Originally Posted by rick-slo

? Restate the question.

Ok, I have a 5.1 speaker system for my computer. When I play a ripped cd through Window Media Player (wmp) it sounds very good and balanced. When I played a mp3 with wmp and recorded it with my set up through Audacity, a bit of bass was introduced. Any idea why?

[rick-slo]

Originally Posted by TBman

Ok, I have a 5.1 speaker system for my computer. When I play a ripped cd through Window Media Player (wmp) it sounds very good and balanced. When I played a mp3 with wmp and recorded it with my set up through Audacity, a bit of bass was introduced. Any idea why?

? You played a mp3 through your speaker system and recording the speaker output with your recording gear? If so, why?

[runamuck]

Originally Posted by TBman

Ok, I have a 5.1 speaker system for my computer. When I play a ripped cd through Window Media Player (wmp) it sounds very good and balanced. When I played a mp3 with wmp and recorded it with my set up through Audacity, a bit of bass was introduced. Any idea why?

mp3s are not 5.1. Is your "ripped CD"?

[Rudy4]

This whole process sounds like its a comparing "apples to oranges" thing. I've followed some similar topics and it turned out to be from the user being unaware that they were recorded at entirely different bitrates.

I'm sometimes surprised by users who state that they hate the sound of mp3s, but when asked about bitrate settings they have no idea that they need to change the default settings of the recorder if they want mp3s to sound anything like a CD.

Not saying that's the case here, but you can never assume too much when a question is posted on a forum.

Your "bit of introduced bass" may be a lack of upper end definition resulting from lower bitrate recording settings. Are you using 320kbps?

[KevWind]

Originally Posted by TBman

I had found tonight that I had changed the input recording to 96000hz and the bits to 32 in Audacity. Turns out this is not a good idea on my system as 44100 and 24 bits sound better. I was going through a test of playback of a mp3 of Al Petteway's playing and recording it, to see what's up with my recording gear when I found this. I also noticed that my recording gear (two condensors, a mxl 991 and AKG perception 150 as a spaced pair into a usb tascam us-122) introduces a bit more bass than what appears in the original mp3. Any idea what could be causing this? Cheap mics, cheap Tascam? The room?

You have combined three completely different discussions with three different sets variables. Into one question.

First off: digital bits and samples and conversion is an entire field of discussion and science in an of itself.
If 24 bit 44.1 Hz "sounds better" than 32 bit 96 kHz ( on the same file, with no other additional variables introduced) then it is function of a deficiency in your system (probably the converters) If you changed something then there is no possible way to accurately compare any difference between the two formats

Second: yes everything ( EVERYTHING) in your signal chain, as well as the room in any acoustic recording, will affect the sound for better or worse depending on many many variables .


Third: acoustic re-recording would be expected to sound different, it is almost virtually guaranteed to sound different .

And as an observation and general rule of thumb:

Using a computer 5.1 surround system to monitor stereo recording is probably counterproductive.

Using a computer speaker system to monitor sound is marginal at best.

Computer sound card conversion is marginal at best

The goal in accurate monitoring and comparison is to reduce the number of variable not increase them

Down converting a CD to MP3 is increasing variables
Re- recording a CD by playing out through a speaker system is increasing variables.
Doing both is exponentially increasing variables.

[MikeBmusic]

You're getting "more bass" because of the room you're doing this in (not because of any bit rate change). I'm guessing you have no acoustical treatment in the room (bass traps) and the bass is getting bounced around the room and the mics picked that up. When you walk around the room listening while playing back the music, you will no doubt hear the bass louder in some places than others, same with other frequencies, these are room nodes.

[TBman]

Great info guys, thanks!

[Andy Howell]

I always record 24/96 but in all honesty many people can't hear anything past 24/44. 32 bits seems a waste of disc space and CPU power!

[sdelsolray]

Originally Posted by Andy Howell

I always record 24/96 but in all honesty many people can't hear anything past 24/44. 32 bits seems a waste of disc space and CPU power!

No human can hear past a 44.1k sample rate (which results in a highest frequency of 22.05k Hz), except for those who claim/pretend to have golden ears. Similarly, nearly all near field monitors, even high end ones, can't reproduce even higher frequencies. Bit depth can affect the accuracy and results from internal processing, e.g., many plugins work better with 32 bit sources than, say, 16 bit sources.

[KevWind]

Originally Posted by sdelsolray

No human can hear past a 44.1k sample rate (which results in a highest frequency of 22.05k Hz), except for those who claim/pretend to have golden ears. Similarly, nearly all near field monitors, even high end ones, can't reproduce even higher frequencies. Bit depth can affect the accuracy and results from internal processing, e.g., many plugins work better with 32 bit sources than, say, 16 bit sources.

I agree and would only add:
I am certainly no scientist and only have a basic understanding of the math and physics involved. I have however read a fair amount of material on the subject, it is my understanding that:
While it is true the upper limit of human hearing is aprox. 20k making 44.1 samples theoretically capable of handling everything in that range . There are a few further considerations involved.
One being at 44.1 k there is very little wiggle room for the converters handling of intermodulation and aliasing distortion accurately before these may affect on the audible range . Whereas at higher rates it becomes less of issue or even a non issue.

Additionally as far as plugins ..Many (depending on the actual digital process involved ) plugins are more easily programed to be more accurate at sample rates above 44.1 which would seem to make 48k or 88k the more desireable choice . However most well programed plugins now internally oversample to achieve more accuracy.
Which would then seem to then make 44.1 the logical choice....

BUT another consideration involved and arguably the more significant consideration than sample rate, is the accuracy of the converter and it's filters.
And perhaps I am stating the obvious but I think it bears repeating --The good news is that converters are getting better and less expensive. The bad news is the concept of "you get what you pay for " does not magically dissolve in the digital realm

Here is an article on the subject.

http://www.trustmeimascientist.com/2...-when-it-isnt/

And here is one paper by Dan Lavery that is referenced in the above article

http://www.lavryengineering.com/pdfs...lity_audio.pdf

[rick-slo]

Very confusing thread discussing bit depth when it seems the case is that the OP played a mp3 through his stereo speaker system and recorded the output of the speakers.
Then the OP compared that recording to the original mp3 and is wondering why they sound different. Well, of course they sound different. It's nothing to do with bit depth.

[KevWind]

Originally Posted by rick-slo

Very confusing thread discussing bit depth when it seems the case is that the OP played a mp3 through his stereo speaker system and recorded the output of the speakers.
Then the OP compared that recording to the original mp3 and is wondering why they sound different. Well, of course they sound different. It's nothing to do with bit depth.

Very true and was more or less pointed out in the first 8 or so replies. But this an internet forum where "thread drift" seems to be an inherent form of "distortion'