Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[dobster]

I know alot of topics are discussed here especially SRC, but for some reason the solutions seem nebulous. Maybe its all the back and forth or maybe I'm just an idiot.

Anyways, what I gathered is and started doing is, record at 44.1 Khz.
Then for the mixing and dsp processings SRC (using quality SRC like r8brain, etc.) to 88.
Then last SRC back to 44.1 for CD.

How does this sound? Do I have the basic understanding down in a nutshell?

anybody

[dobster]

by the way, after reading even more material especially Dan Lavry's, I wanted to add then that my converters are the delta 1010. because it seems it came down to the statement "if the converters are good enough then samle rate doesnt really much matter after 44.1khz"
is this true???

and if it is, i'm still stuck with the question is it better to upsample for the sake of processing, esp. non-linear processing?
if so, whats the best way to go about upsampling when the tracking is all done and you;re looking at the tracks sitting in the DAW?

[Gunnar Hellquist]

dobster wrote on Sat, 12 August 2006 21:25

Anyways, what I gathered is and started doing is, record at 44.1 Khz. Then for the mixing and dsp processings SRC (using quality SRC like r8brain, etc.) to 88. Then last SRC back to 44.1 for CD. How does this sound?

Well, if it sounds good for your ears then by all means let them decide. All recording and processing is about making things sounding good. If it sounds good it is good.

What my ears tell me is that all converters sound a little different at the speeds they offer. Some are quite different, some not much at all. Same goes with processing, some plugins seems to sound quite different at different speeds. Not necessarily better at a higher speed though.

So what you need to do is to test your equipment and your plugs and use your material to let your ears decide. Personally I run everything at 44.1kHz and have no second thoughts.

Gunnar

[dobster]

Hey Gunnar,
do you do any non-linear processing? if so, you still remain at 44.1?

[Gunnar Hellquist]

Yes, I run processing at 44.1 as well. I use Samplitude which to my ears has good quality effects. The built-in EQ, compressor / limiter and impulse reverb is what I tend to use.

Gunnar

[dobster]

hmmm, I'm definitely not doubting the quality of the plugins - to be honest, I haven't used them so I don't know either way. But, it seems that no matter the quality of whatever plugin, there's a universal themse to them all when it comes to non-linear processes. and the cure, which is what i think i'm grasping is oversampling for non-linearities

[Gunnar Hellquist]

Well,
  there is theory and there is practice. Theory is when things does not work, but you know exactly why. Practice is when things work, but you have no clue as to why. Often theory and practice is combined so that nothing works, and we have no clue to why.

When it comes to sound however, we do have the very best tools ever, our ears. Using them we can skip a lot of the theory and simply go straight for practice. So, as I said in an earlier post, if it sounds good it is.

Gunnar

PS: theory does have its place. It is a very important tool and often the only way to take the really large steps. But as some great thinker said: science is the ugly, messy meeting where reality shatters even the most beautiful theories.

[danlavry]

dobster wrote on Mon, 14 August 2006 07:46

hmmm, I'm definitely not doubting the quality of the plugins - to be honest, I haven't used them so I don't know either way. But, it seems that no matter the quality of whatever plugin, there's a universal themse to them all when it comes to non-linear processes. and the cure, which is what i think i'm grasping is oversampling for non-linearities

The non linear process generates all sorts of energy. Say we want to eliminate all the energy over -100dBFS. In an analog system, we could filter unwanted energy, but with digital the energy aliases as it gets generated so it can not be filtered.

So the idea of oversampling moves Nyquist (the aliasing point) up, and the hope is that when Nyquist is higher, by the time you get there, the unwanted energy will be lower energy. The lower energy will fold back to the audio, but at lower levels.

All that is great, but one needs to QUANTIFY it. In general, a non linear generation of energy goes down in amplitude as you raise the frequencies, but it goes down very slowly with frequency. If you look with a scope, or compute the numbers (for a given non linearity), you may be wise to recall that while 1% deviation on a scope picture looks near perfect, the ear is near logarithmic so 1% is only -40dBFS, thus rather disturbing, especially when if folds into say 1-3KHz...

So the solution may be upsampling, but in practice, you may need to be ready to upsample to 10MHz! Then you do the process, and follow it by a huge down sample ratio.

Digital and non linearity do not go hand in hand. )Digital and linear processing do go hand in hand).

Regards
Dan Lavry
http://www.lavryengineering.com

[blueintheface]

Gunnar Hellquist wrote on Mon, 14 August 2006 16:36

Well,   When it comes to sound however, we do have the very best tools ever, our ears. Using them we can skip a lot of the theory and simply go straight for practice. So, as I said in an earlier post, if it sounds good it is.

Ears are fine, but often the brain gets in the way . . .

[danlavry]

Gunnar Hellquist wrote on Mon, 14 August 2006 16:36

Well,   there is theory and there is practice. Theory is when things does not work, but you know exactly why. Practice is when things work, but you have no clue as to why. Often theory and practice is combined so that nothing works, and we have no clue to why. When it comes to sound however, we do have the very best tools ever, our ears. Using them we can skip a lot of the theory and simply go straight for practice. So, as I said in an earlier post, if it sounds good it is. Gunnar PS: theory does have its place. It is a very important tool and often the only way to take the really large steps. But as some great thinker said: science is the ugly, messy meeting where reality shatters even the most beautiful theories.

I do not share with you the definitions of what theory is, and what practice is. Your description of "how Often theory and practice is combined so that nothing works, and we have no clue to why" seem to suggest that you had some dealing with incompetent people. A competent person that is well equipped with theory and experience is likely to do good.

Yes, the ear is the "final judge" of what sounds good to you, and someone else may disagree. That is because tastes are subjective. That is why we do not get in this forum into "back and forth" about what sounds good or bad. Please look up the first post in the forum under "introduction".

The ear can tell you what sounds good to you, but the ear is NOT an ohm meter, it is not a soldering iron, it is not a scope... Say you hear a differance between 2 sounds. I can not ever doubt it. But do you know if the difference is due to some 30KHz energy? Or due to some dielectric absorption of some cap? or due to a certain type jitter? As a rule, you do not!

The ear is the goal, and yes it is the most important, because it is the end result. And yes, you can train the ear to "measure" frequency, but not very accurately. I do not know anyone that can tell a 1 cent pitch deviation, not even while switching back and forth. That is about 590 parts per million. A cheap frequency counter will do 10-100 times better!

There are huge number of variables that go into a design. They all need to be set right. The chances to do it right by ear are lower then winning the lottery 10 times in a row.

So again, you can use your ear to decide what to use, what you like, what you do not. But there is no need to put down the theory, which represents over 100 years of millions of electronics engineers careers achieving "inch by inch" progress to get us where we are today (which is not perfect, yet a huge advancement).

Regards
Dan Lavry
http://www.lavryengineering.com

[Transcending Music]

well put Dan...

[Gunnar Hellquist]

Dan,
 I agree completely with you. Writing in a foreign language on the net and my intended irony did not come through. I am sorry for that.

I firmly believe in understanding the theoretical concepts behind what we do in a DAW, but also to be very careful to avoid confusing my incomplete picture of the theory with reality. It is very easy for me as user to take only part of the theory and forget to check it against real-world results. As in the discussion that started it: is it better from a theoretical point to upsample to 88.2kHz in order do processing? It surely must depend on a lot of real-world factors and at the end of the day it is the sound that counts. One factor is that every piece of equipment and every plugin and every SRC algorithm is different. So I as a user simply have to try the result with my own equipment and using my ears.

[Ronny]

blueintheface wrote on Mon, 14 August 2006 22:12

Gunnar Hellquist wrote on Mon, 14 August 2006 16:36 Well,   When it comes to sound however, we do have the very best tools ever, our ears. Using them we can skip a lot of the theory and simply go straight for practice. So, as I said in an earlier post, if it sounds good it is. Ears are fine, but often the brain gets in the way . . .

Well said, the ears are great devices, but the auditory cortex' ability to assemble that information correctly everytime and the humans ability to recall from memory exactly everytime, lacks way behind the design of the ear. There are a host of human sway elements involved, and why it's good practice to confirm what you think your ears are telling you, with the scientific measurements and the blind and double blind tests.

If you hear a difference between two samples, yet they totally phase out in a null test, which are you going to beleive, your ears or the science? I'll take the science, because although I'm pretty sure my ears are good, I know that the auditory cortex can be altered from hour to hour by mood, climate, substances such as too much caffeine, over exposure to sound (ear fatique), sensory inhibition (a survival trait instinctive to humans), sighted evalution (expectations leading to imagination)and cross-sensing. A non-audio example of how our senses interact is tasting food. When you have a cold and clogged sinuses, food tends to taste bland and all dishes the same. The reason is that the olfactory glands play a key role with the taste buds in determining flavor, it's not just the taste buds. ITR, the taste buds and nose are dependent on one another to evaluate food. Same applies to the sighted gear evaluation, the eyes affect the ears. That's why the blind test is important and typically gives unexpected results. Often it brings more questions than it answers, but it's the best we have for eliminating the human sway factors.