I need some clarification from someone who knows what’s going on under the hood of this very capable program. I’m kind of new to Cubase as a primary DAW so thanks in advance if this is too basic a question. Does Cubase 6 use 32 bit floating point internal processing for operations such as changing gain of an event and crossfades regardless of the bit depth settings in the project set up page?
I track using 24bit converters so I assumed that was the bit depth I should choose but I wonder now if I should be at 32bit for the best fidelity when I’m mixing or even just changing gain. My computer is very capable and pimped out with ram etc. so I’m more concerned with fidelity than conserving cpu resources.
Thanks, I’m sorry if I sound dense here, but when you say it processes at the max are you saying it processes at 32 bit floating regardless of the bit depth setting in the project page?
While I’m asking pesky questions, what’s the upside of saving files as 32 bit floating?
Yes internal resolution is 32 bit even if you record at 24bit.
32bit float isn’t worth bothering with in my opinion…though I’m sure some will disagree.
FWIW I believe a lot of people (certainly myself in the past) tend to question the quality of recording or playback or mix engine etc when actually it’s their inexperience in mixing that leads to poor quality output. Poor tracking with cheap mics & lots of EQ boosts in a mix will soon make a track sound crappy & what people might perceive as digital.
Not saying this is you though…sorry if that’s drifted a bit OT.
Thanks Grim. You make a good point, but according to those that know much more than me, even the most pristine tracks and mixes can be degraded by truncation and other bad objective stuff that can happen deep in the number crunching phase of a DAW or digital device.
It’s pretty much always “garbage in - garbage out”, but sometimes you get garbage out even if you insert excellent ingredients. This is certainly not the case with Cubase but I have heard some pretty unusable digital mixers and a/ds for example.
Some of the people I work with come from a classical music world and can be very fussy about sonic quality as they are accustomed to a lot realism (from some pretty high end DAWs) so I like to be sure I’m not doing anything destructive to the signal if I can avoid it.
Just stick with 24 bit! I like to record at 48hz 24 bit and get good results…if you realy wana go 32 do a test on yourself make two recordings one 24 and the other 32 mix them down chuck them onto cd or whatever and listen to both and without looking what one is playing see if you can tell the difference…i dout you can.then make your own mind up.
I’m also new to Cubase and I was wondering the same thing as the OP. I come from SONAR, where you can set up the bit depth for different tasks (i.e. recording, bouncing and playback). As far as Cubase goes, I’ve only seen an option for recording under the project options, but I don’t know what is going on under the hood. Is Cubase using the project’s bit depth as the resolution for operations like exporting/bouncing/freezing? Or is the audio engine always processing at 32 bit float, including exporting/bouncing/freezing? The latter would make more sense, if I’d had to guess, but I wanna know for sure. Where can we find factual information about this?
Thanks fellow members for all your help. In the interest of putting this on the record for fellow users, here’s a verbatim response I got back from Cubase customer support:
“We would recommend using 32 bit floating point for the file types in the project so that you take maximum advantage of Cubase’s processing and mixing capabilities. This allows the any changes in the gain, plugins, mixing, panning, etc to be processed in their highest bit depth fidelity. I hope this information helps.”
I did specify that my primary concern was maintaining the highest fidelity over conserving resources and hard disc space.
What I take away from all this is that there is a concrete (if sometimes subtle) difference in the preserved audio quality of processed files. Much like making the case for using higher sample rates, it may or may not be worth the “cost” to a particular user in a particular application.
Hmmm, he’s saying to use the 32 bit float option for the file type, but is this the same as the Project’s Bit Depth setting? Since there is no mention of the Project’s Bit Depth in his reply, it makes me wonder if perhaps there is another setting somewhere in Cubase. Unfortunately, I’m currently not in front of my DAW to confirm this. I think I’m more confused now
All internal processing is done at 32fp regardless of source bit depth, unless performing a process that writes to disk like bounce or audio processing when the project bit depth is observed.
I used to set this to 24bit and switch to 32bit once I’d finished tracking but as I’m way to lazy to do this \i now just set the project bitdepth to 32bit and don’t have to worry any more. It doesn’t incur any CPU overhead and uses up a bit more disk space but disk space is cheap so who cares.
That’s exactly what I was pondering throughout the day after my last post, that perhaps we need to set the project’s bit depth to 24 bit until after recording and then set it to 32 float afterwards for bouncing/freezing. I don’t know that I’ll like that, having been accustomed to setting and forgetting these settings in the past. But, at least Cubase makes this really easy, since you only have to click on the info bar (I think that’s what it’s called) in order to recall the Project’s Setup window. Not that big of a deal either unless I bounce, which I don’t do as much these days. As long as all the processing is being done in 32 bit float, eveything else is good.
I did some more reading up…there are plenty of discussions of the benefits or not of using 32bit float online.
My conclusion is that while there is certainly no harm in using 32bit (other than additional HD load & filesize)) I’m also confident that I’m not losing any quality at all by mixing at 24bit…even when bouncing or offline processing it seems it would take a large amount of multiple processing before there was any appreciable difference.
so that you take maximum advantage of Cubase’s processing and mixing capabilities. This allows the any changes in the gain, plugins, mixing, panning, etc to be processed in their highest bit depth fidelity.
Doesn’t the internal 32bit processing take care of ALL the mentioned processes so they are always done at the highest bit depth fidelity?
Yeah I thought that too, I would think that support either doesn’t understand the question or thought the explanation was too complicated for us “dumb” users
I think the 32bit fp off line would have to be applied multiple times to see any real benefit from 24 bit, but none the less it does use the project bit depth setting and for the pedantic (of which I can be guilty) setting the project to 32fp yields the more accurate results albeit at the expense of some disk space.
