Question about bit depth settings in X1c...

Hey all.  I've got a Focusrite Pro 24DSP interface which I had been running at 44.1.  I bumped it up in both X1 and the Focusrite mix control to 48 and 24 bit depth and was getting slightly better latency.  I'm also running the 24DSP at 96samples.  But I noticed in X1's control bar it was reading 48 Khz but 16 bit depth.  I went into my audio settings and even tho it shows 24 in the bit depth window it's greyed out and can't be changed.  There's nothing I can find on my 24DSP that allows me to change "bit depth" that I can find.  Should I not even worry about this or is something amiss??
 
PS: Just installed the latest drivers for the 24DSP...no change.
 

Hi olpal,
Go in to Preferences and set the record bit depth to 24.

best regards,
mike

mike_mccue


Hi olpal,
Go in to Preferences and set the record bit depth to 24.

best regards,
mike

I suppose its good duplicate.

Im set to 24b here . no burps farts or stuttering...

OK, I did that and then rebooted and now it's showing 48 and 24 on the Control Bar.  I guess my question is...why couldn't I change it in the audio dialog and why did I have to go to the file dialog???

Ah Ha.  Thanks, John, ol pal!  And what, if anything (ducks head) will I gain from switching from 44.1/16 to 48/24???
 
Again...I just thought I'd try it since I'm so high falutin 64 bit now:-)
 

What would you gain OlPal? Not a whole lot except 24 bits is a much better record bit depth on two levels. First is the low noise it offers and second the greater dynamics it has.  To me recording in 16 bits requires a lot of work in keeping the level high without clipping. 24 bits is much more forgiving. You can have it recording lower in level and still have a quiet recording. Not so with 16 bits. 

One, I haven't had any clicks show up in my waves since going 32bit. At 24 I used to get individual samples that would go wacky, especially in long tracks, and make a clicking sound. I'd have to open up the track in Sound Forge 9.0 to fix it. Since I've switched to 32bit I haven't had to open SF once to edit. I now use it as the worlds most expensive MP3 maker. :) Second, I haven't had any clipping problems, even when I drive the signal in to the red. I never intentionally do that, but it does happen from time to time and so far I haven't had to scrap a take because of it like I had to when running 24bit. Third, HDD space isn't an issue anymore so the extra space isn't a concern. And fourth, Sonar supports it (and 64bit depth as well) so I take advantage of it.

Bub that is due to clipping. Your levels were way to high. 24 bits is an integer format that although it has a lot of "headroom" a 32 bit file is also floating point with that much more headroom.

Really the only valid reason to go 32 bits is for use in another DAW and even I don't see the real advantage.

So why not do everything at 64 bits? Some swear they can hear a difference but then so do those that use a high sample rate.

If you are aiming at placing this work on CD the 24 bit limit is a great compromise. Going from it to 16 bits is not that much of a change as compared to 32 bits or 64 bits. Truncation should not be so severe. Plus one saves on HD space.

All this have been discussed so much that its one of the most debated subjects on this forum. 

John

Bub


I prefer 32 bit depth personally.

My current settings ...

M-Audio Fast Track Ultra - 64 Samples (The latest drivers don't support it. Had to revert to old drivers but all seems well.)
Sample Rate - 96kHz
Bit Depth - 32 bits
RTL - 3.2ms

I had a problem one time where no matter what I set the bit depth too it was froze at 16. I ended up deleting the AUD.INI file (IIRC) and let Sonar rebuild it. I think it was the aud.ini ... can't remember fer shur.

Why do you prefer 32 bits? You do know you can not hear that bit depth.

We have had many disscussions about this, the benifits and this needs to be proved is if you have a 32 bit audio file and everything is being processed at 32 bit, it is less taxing on the system.  So the theory is if you have a 24 bit audio file, the DAW has to process that file at 32 bit so instead of the CPU being used for effects and instruments, it's being used for upsampling (not too sure on the teminology) it becomes a little la la when you consider the power and thruput of a modern computer but I can see the difference the whole sysytem does run better.  I use 32 bit for all my new recordings.
 
There is no audio quality difference, try this: clip an instrument track and freeze it at 24 bit and 32 bit, you can't get internal digital distortion at 32 bit but at 24 bit there is def less headroom.
 
Finally if you have recorded a project at 24 bit set everything to 24 bit, I dont't really know why I do this but i just believe that all settings should match.
 
Any questions??
 
Peace Ben

BenMMusTech

John

Bub


I prefer 32 bit depth personally.

My current settings ...

M-Audio Fast Track Ultra - 64 Samples (The latest drivers don't support it. Had to revert to old drivers but all seems well.)
Sample Rate - 96kHz
Bit Depth - 32 bits
RTL - 3.2ms

I had a problem one time where no matter what I set the bit depth too it was froze at 16. I ended up deleting the AUD.INI file (IIRC) and let Sonar rebuild it. I think it was the aud.ini ... can't remember fer shur.

Why do you prefer 32 bits? You do know you can not hear that bit depth.

We have had many disscussions about this, the benifits and this needs to be proved is if you have a 32 bit audio file and everything is being processed at 32 bit, it is less taxing on the system.  So the theory is if you have a 24 bit audio file, the DAW has to process that file at 32 bit so instead of the CPU being used for effects and instruments, it's being used for upsampling (not too sure on the teminology) it becomes a little la la when you consider the power and thruput of a modern computer but I can see the difference the whole sysytem does run better.  I use 32 bit for all my new recordings.
 
There is no audio quality difference, try this: clip an instrument track and freeze it at 24 bit and 32 bit, you can't get internal digital distortion at 32 bit but at 24 bit there is def less headroom.
 
Finally if you have recorded a project at 24 bit set everything to 24 bit, I dont't really know why I do this but i just believe that all settings should match.
 
Any questions??
 
Peace Ben

My audio engine is set to 64 bits.  I still record at 24 bits. Mainly because that is the best my hardware can go. If I record at 32 bits a lot of it is empty with padding. No real data is being recorded to the extra bits. No matter what Sonar will need to down grade the bit depth to monitor what ever it is you are playing or recording. It can't output at 32 bits. Only the files can be 32 bits or higher. That is not an issue for processing but file making.

I hope I am making myself clear here. There is no way to avoid 24 bit or lower processing. It is not done in the hardware or the hardware driver its done in Sonar so you and anyone else can hear the music. Otherwise you will have the most pristine sound that is never heard.

BenMMusTech


Thats the point though John, the extra padding as you put it helps the internal system run better, or so the theory goes, if you mix in 64bit floating point, so should your audio files be 64 bit.  Yes your audio interface can only handle 24 bits but the downsampling is handled by your interface and not your CPU/Sonar.  This is how I understand it (correct me if I am wrong).  So theoretically you computer runs better

Ben

That is nonsense. Processing has to take place no matter what bit depth you have the internal processing set to. And it has to be processed to 24 bits or lower for you to hear it. How on earth do you think what you hear is getting out? Further since when did larger numbers become easier to process than smaller numbers for a calculator? Since when did floating point numbers become faster to process than integer numbers?

In the past I kept my internal audio engine at 32 bits in Sonar because 64 bits was too much of a strain on the CPU. That is no longer the case because I have a much stronger CPU. That shows just how off base you are here.

Maybe its the 32 bit file making and the 96 Khz sample rate that is your problem. LOL


BTW I thought you couldn't get X1 to work.

John

BTW I thought you couldn't get X1 to work.

It's not when I use Guitar Rig and a few other plug-in's, but that doesn't stop me from rendering.

Guitarpima

I read something somewhere where 88khz is better than 96khz. It had something to do with the math being easier for the downsampling.

I saw that somewhere too. Then in another thread I saw it's better to use 96kHz for guitar plug-in's like Guitar Rig because it makes them sound better.

... No weird artifacts since going 32bit.

+1

Here's my take on it all ... do some tests on your own system and see what sounds and works the best for you. There's no right or wrong way, just a lot of options ... which is a good thing. For example, 44.1kHz sounds better than 96khz on my Fast Track Ultra ... but I can't live with the latency at 44.1kHz that's why I switched back to 96kHz. I really think it all depends on your system and the mix of components you are using.

I'll give ya's my take. I've done so many tests on this, it was like chasing guitar tone. I'm telling you, I must have done this at least 100 times or more. Here's what I've come up with using (in my opinion) 2 of the best soundcards out there with killer converters in them. That said, 2 really good cards is my problem. The better your card, the less you will notice these differences because the pro cards are consistent with all bit and sample rates. The cheaper cards....you will definitely notice a difference on because well...the converters are pretty bad on some of them.

16/44 vs 24/48: Not a night and day difference to most people, but it depends on how good or bad your card is. To me, the difference between these 2 is 16/44 is a one inch tape machine at 15ips, 24/48 is a 2 inch tape machine at 30ips. That's what I hear using both a Layla 24/96 card and an RME Fireface 800.

24/88 and 24/96....no change, no difference and what I believe to be totally useless for what most of us do here. Rock, metal, pop, dance, techno, blues etc and anything that has a majority of direct sound instruments, sims, samples etc....useless, waste of hard drive space as well as pc resources.

Now, if you are going to record a choir, orchestra, Jazz ensemble or something that is using loads of clean signals/sounds, mics, room sounds and ambience, you WILL notice a difference using 24/48/96 settings. That I promise you. But if you did a rock project at 24/48 and then re-did the same project from scratch using 24/88 or 24/96, you won't notice a difference. Don't even kid yourself there. Do it and see....but here's the catch. Create 2 mixes. A mix at 24/48 then one at 24/88 or 24/96. Name them mix 1, mix 2...and then grab one of them...whichever you decide, copy it and rename it mix 3. Make a playlist or something that will start playing at random and let your wife, girlfriend or one of your friends initiate it. Write down which you think is which as it plays and make sure the person running this is keeping track of which files are playing as well. The reason for the duplication of one of them is to throw you off. Or, just tell them to play the files at random.

Now, the other key to this is to make sure when you perform the instrumentations on these 2 projects, that you do not make any obvious mistakes or errors/artifacts within the mixes that can tell you which mix you are listening to. For example, if you made a mistake for a split second on the guitar track in the 24/48 mix that isn't on the 24/96 mix, you'll know in an instant which you are listening to.

32 bit float: I personally feel this is a load of BS. I know about the math and all that stuff...but I don't buy it. If I can't hear a difference that literally makes a difference, it doesn't exist. Soundcards do not play audio at 32 bits. When you process, bounce-down or apply effects etc, some joker claimed "you don't lose anything at 32". Well, if I lose something in 24, it's undetectable to the human ear...therefore, it's not something I will waste any system resources on.

64 bit processing/mix engine and anything 64 that isn't Sonar 64 the program: No difference whatsoever....plus it gives me artifacts just about always. When it does work, there is no difference that leaps out at me and shakes me to use it. Remember....and this is important. There are going to be several jokers in the DAW world that will attempt to show you math, graphs, charts and other horsesh!t that attempts to prove a point. If you can't hear it...no one else will either, therefore, it's hype and doesn't exist. If I didn't test this so extensively over the years, I'd not mention a word here. If something doesn't totally grab you that makes a huge difference to where you say "oh yeah, now THAT makes a difference and sounds 100 times better than the other" don't waste the time...no matter what it is.

I read about people doing tests using high end pre's and other stuff that you supposedly can't live without...most of that stuff makes a 5% or less difference. And some of that stuff you can hear. For those that claim they can hear differences with this stuff....how much of a difference do you think it is making....2% if you're lucky?

24/48 for instruments with a bit of dirt going on is all you need. 24/96 if you have a super clean front end as well as super clean sounds going in. That means....non-distorted guitars and nothing with drive/dirt. That's where you hear the differences in the sample rates. Samplers, dirty guitars, over-driven bass, B-3's with more drive than normal....all that stuff with drive....you've degraded already by adding drive....which will not change the sonic sounds these things put out using a higher sample rate. I'm not perfect in anything I do...but ears I got. They never fail me....but for the life of me, I can't hear some of this stuff. Try it for yourself though and see. Also remember....if your soundcard is professional grade card, you will notice less differences because that's how they are supposed to work. Good cards deliver the goods at all bit and sample rates. On lower quality cards, most people can hear a difference because well...they ain't made for pro recording. :) If my comments here conflict with anyone, I'm just telling it how I've lived it. If these other settings work for you and it's a huge difference, you're right where you need to be.

-Danny