Recording Sample-Rate? (Technophiles Help!)

S

Somelsewhere

New member
Alright, so I recently downloaded the Reaper recording software, and so far, I have to say I am more than pleased.
But I can't help but feel that I'm not capitalizing on the "oomph" of it.
So I went into my audio preferences and found that my sample format was running at 16-bit, so I naturally brought that up to 24-bit (as that is the capacity of my Lambda usb-interface.)

Then,
I saw that the "sample-rate" was set at 44100 Hz,
and the "buffers" at 8x1024 samples.
(the latency is shown at 185 ms with these settings.)

Now, I don't profess to be very computer savvy, hell, I'm just going to admit right now I don't even know what a sample-rate is...

So can anyone help me understand/ give me a suggestion to tweak these settings so that I can get an even better recording out of an exemplary piece of software? (loving Reaper. Simple, but complicated if you want it to be.)

My computer has an Intel Core Duo T6500, 4GB SDRAM, 320GB HDD, Vista, and I'm running the 64-bit version of Reaper. (aka the newest v. I guess.)
 
Don't know anything about the Lambda but don't care for
USB or FW for that matter.

You should have no problems with setting it 24bits and
48KHz. I normally only use 2 256K buffers and have a latency
of 2.3ms. You should be able to get yours to to less
than 10ms even with USB.
 
What do you think the best quality would be though? My computer probably wouldn't have a problem doing a lot of things, but I want to make this baby work hard for me.

So you are suggesting I change the sample rate from 44100Hz to 48000Hz,
and the buffer rate from its current 8 x 1024,
to 2 x 256?
(it's completely possible and likely that I misunderstood.)

Would I benefit even further from increasing the sample rate from 48khz, to 96khz,
or does the limiting factor of my Lambda interface (24 bit) make such a change useless?
(also, is latency in the ms range something that really affects sound quality?)

I suppose I might even be asking for a crash course,
and anyone willing to answer these questions,
or even to educate me,
will have my most earnest, noobish, thanks.
 
44.1khz is the sample rate used in CD's while 48khz is the sample rate used in videos. I would recommend staying with 44.1khz and 24 bit. In the end, it doesn't really matter which you choose. Most software will convert from one to another.

There are people who will argue that they can hear a difference between 44.1khz and 96khz. But for real, in most home recording scenarios, there's going to be a lot of other factors that will oveshadow the audible difference between 96k and 44.1k. Running with 44.1khz yields a significant savings in cpu resources over 96khz. I used to run with 24/96k and my cpu load was about 60% for about 10 tracks with plugs. I have recently switched to 44.1khz, don't notice any audible difference, yet my cpu load barely even registers.

As for buffer and latency, if you're monitoring your recorded track directly and in realtime (ie, not going to the computer first, then coming back to your interface) then set a high value for your buffer. This lessens the load on the cpu. If you're monitoring through your DAW, then a lower setting will give less latency. Anything below 10ms shouldn't give you that annoying echo or delayed effect.

Hope this helps
 
The audio system was preset at "WaveOut",

should I switch it to the ASIO option?

(the others being WDM Kernel streaming, Direct sound, ASIO, dummy audio)

I'm assuming the WaveOut is the only compatible format for my Lambda input,
but given that I don't even understand the concept of an "audio system", any gentle corrections would be appreciated.
 
pipelineaudio said:
He seems not to be using asio too, thats a cpu chomp if not
Click to expand...


I guess I missed that part. You want to use ASIO whenever you can. I'm not familiar with the Lambda interface, but ASIO drivers automatically set the optimum buffer and latency settings. They also provide the most efficient data handling for the audio signal.

Here's a quick, but good write-up.
 
Lock it on 24bit/44.1khz ASIO and forgetaboutit.....

Only reason to go 48k is if you are using dedicated equipment (video, ADAT or cards like SBLive that are locked at 48k.)

You will have to render/export down to 16bit/44.1k to make CDs.
 
Although the difference between say 44 and 96khz is hard to discern...the difference between recording in 16 and 24bit is much more pronounced...is that right? I heard it was particularly true for things like acoustic guitar.
 
Monkey Allen said:
Although the difference between say 44 and 96khz is hard to discern...the difference between recording in 16 and 24bit is much more pronounced...is that right? I heard it was particularly true for things like acoustic guitar.
Click to expand...


I've never done any comparisons myself, but yeah, that's what I've heard also. I'm no expert and can only repeat what I've heard/read: the difference between 16 bit and 24 bit is more noticeable when you start adding processes to the signal. Going through converters, plugs, more plugs, exporting, changing formats, etc., then mixdown multitracks to stereo, whatever it is that happens to the audio signal, each stage adds a little bit of error. If you have more bits to represent the same data, those processes can perform the math more precisely and induce less error. In the end, you'll have a truer repesentation of the audio.

That's about the best that I can explain it because I'm not sure I fully understand it all myself. :o
 
You must log in or register to reply here.
Back
Top