Page 1 of 2

Console and tape emulation at 88.2--is it worth it?

Posted: Wed Jun 06, 2012 6:33 pm
by justinmorell
Hi all,

Just looking for some opinions. I have a session at 88.2kHz and would so like to use some of the AlexB and CDSM emulations. I'm wondering, is it worth the sonic price of SRC (I have izotope) to upsample to 96kHz to be able to use the emulations?

Thanks,

Justin

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Wed Jun 06, 2012 7:58 pm
by biomuse
Oh gosh I would think that the difference between 88.2 and 96 would not be worth it at all. It's all getting downsampled to 44.1 at the end anyhow, I assume? You'll get 99.9% of the goodies you're after just letting Nebula downsample to 88.2 (which it does well). Not worth it.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Wed Jun 06, 2012 8:34 pm
by justinmorell
Just so I'm clear on your point, you think the sample rate conversion (whether it's from within Nebula or done elsewhere) is not going to have much effect on the quality, and that whatever loss there may be in the SRC is worth accepting in order to get the benefit of the emulations?

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Wed Jun 06, 2012 9:20 pm
by Martinez
Hi there Justin, it's always worth it sonically as far as i'm concerned.
Just so you know, it is not a case of it's all going to end up a 44.1 anyway, it's a case of once all the audio files a mixed together at 96khz and sound better so then will they sound better at 44.1khz.

The sonic benefit of high sample rates stack across multiple tracks, so the more tracks the more benefit it will be.
When you do a mix and bounce it or record it live to your hard disk the final product is a snap shot of the over all mix and is not magically summed together after it leaves your daw.

In short, yes it's totally worth it providing you have a decent app to do the conversion which from what I hear you do.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Wed Jun 06, 2012 9:22 pm
by david1103
i would say the SRC makes a big difference in quality for nebula, at least in the nebula pre-amps i tested with downsampling from 96k to 44.1. can't confirm for 88.2 but i would imagine it would be the same.

i would upsample to 96k and run your project in that if i were you.

best solution of all is of course to try it both ways if you have the time! if you do, please upload a mix out of interest :D

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Fri Jun 08, 2012 3:21 am
by justinmorell
Thanks. I think I will go ahead and upsample to 96k and work from there. My little CPU will continue to grow more and more angry with me, but that's life. :-)

Can't wait to finally work on a project with all of these emulations!

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Fri Jun 08, 2012 8:05 pm
by biomuse
martinez wrote:Just so you know, it is not a case of it's all going to end up a 44.1 anyway, it's a case of once all the audio files a mixed together at 96khz and sound better so then will they sound better at 44.1khz.
I think you and others have misinformed the gentleman. It's of course true that working at a high sample rate has benefits that translate to the sample reduced master. The point is that any difference between 96 and 88.2 in Nebula is going to be swamped by that last SRC to of the mix to 44.1. It's not as though his session rate is at 44.1 to begin with, in which case the upsampling might make a real difference. Nebula's SRC is more than capable of producing fantastic results with 96->88.2 conversion. I wouldn't bother upsampling everything in this situation.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Fri Jun 08, 2012 10:21 pm
by Barendse
Sample up to 96.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sat Jun 09, 2012 6:47 am
by biomuse
Methinks you have all lost your minds. :shock: Any difference will be perfectly inaudible in a final mix, and you're instructing the OP to go through major inconvenience to get it.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sat Jun 09, 2012 2:58 pm
by ngarjuna
biomuse wrote:Methinks you have all lost your minds. :shock: Any difference will be perfectly inaudible in a final mix, and you're instructing the OP to go through major inconvenience to get it.
+1. People vastly overestimate the effect of SRC.

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sat Jun 09, 2012 3:59 pm
by yr
who do you believe? X that says Nebula src is terrible or Y that says the opposite? -it might be more simple (and much more conclusive) to test it for yourself. After all, nobody on the Acustica forum has your ears...

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sun Jun 10, 2012 8:12 am
by CoolColJ
I don't know about "little difference" - I do know the original at 96k has way more top end, even if I can't hear up there

And any aliasing will be filtered out once you downsample back to 44.1k on mastering, with a good convertor like Iztope RX

Working directly at 44.1k will definitely have more aliasing from everything you do, and oversampling does not sound as good as working at higher rates

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sun Jun 10, 2012 9:51 am
by RJHollins
As to the OP's original question ...

The issue would be ... Can a significant difference be heard between 88.2 and 96k sampling rate !?!?

Next ... if 44.1k is to be the final destination rate,
then we need to consider what every sample rate converter has to deal with.

From even the 'less stellar' converters, the rate change from 88.2 to 44.1 is about as simple as it can get, with minimal conversion induced issues. A 2:1 conversion.

Now when we convert 96 to 44.1, the conversion becomes a 2.17687075 ratio. This has no uncomplicated conversion process.

YES ... many of these hi end converters work well above the source and destination target rates, and involve complex mathematics in which to derive the bit values for the destination.

The bottom line ... there will be a difference ... that is the parameter results of 2 unrelated timing rates.

I'm not privy to the inside engine workings. For myself, I use 'r8brain Pro'. This is considered a good SRC app. Nonetheless, there are differences. Not necessarily deal breakers.

Sample rates are not the only downsizing we must deal with. I would much prefer staying in the 24 bit world, and wouldn't mind either 88.2 or 96k as the final destination.

As engineers, it is priority to maintain the highest quality to whatever media target we must package for.

This is similar to what is required from our mixes ... they must play well with others :)

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sun Jun 10, 2012 11:39 am
by sneaky
Can a significant difference be heard between 88.2 and 96k sampling rate
not to my ears
Can a significant difference be heard between 44.1 and 96k?
Yes. Night and day. Requires a lot of rendering out on my machine but its worth it

Re: Console and tape emulation at 88.2--is it worth it?

Posted: Sun Jun 10, 2012 5:08 pm
by biomuse
RJHollins wrote:From even the 'less stellar' converters, the rate change from 88.2 to 44.1 is about as simple as it can get, with minimal conversion induced issues. A 2:1 conversion.

Now when we convert 96 to 44.1, the conversion becomes a 2.17687075 ratio. This has no uncomplicated conversion process.
sneaky wrote:Can a significant difference be heard between 88.2 and 96k sampling rate?

not to my ears
Can a significant difference be heard between 44.1 and 96k?
Yes. Night and day. Requires a lot of rendering out on my machine but its worth it.
Both correct and, taken together, exactly why instructing someone to upsample their entire 88.2 session (whose final destination will be 44.1) to 96 so that Nebula doesn't have to convert its kernel content from 96 to 88.2 is silly and irresponsible.

Whether the popular "Upsamplit!!!" is the correct answer or not depends on the actual situation.