View Single Post
Old 26th October 2019
Lives for gear

Originally Posted by psycho_monkey View Post
I can’t speak for what you’ve been told, or what you think you’ve been told.

Only for what I understand to be the case.

And if I’VE misunderstood things - someone correct me!

But I’m pretty sure that’s not the case.

It’s not that hard to understand - it does make sense. You just can’t extrapolate to what you think, stuck to what you know.
I do trust you. I agree. And I also agree to ask: please someone else chime in here if there is more to this mystery.. BTW, seriously, who did invent 32bf ?

Originally Posted by psycho_monkey View Post
- what apogee are you thinking is “crappy”? All the apogee products I know are pretty decent.
Ok, I started off decades ago with the Apogee Rosetta (my first converter !)and then later advanced to the 16X's. These are the absolute penuchle for Apogee. I stand behind these. This could be a extremely long story, so I'll keep it simple, and I know I'll get flack here, lol but - Once the 16X's support was dwindling, and I was expanding I did a extremely involved converter test. The right way, blah blah blah - Summary of it all - I ended up with a JCF Latte for Stereo tracking at times when I want that flav and 6 Prism ADA-8XR's for my multiple channel mixing/recording because these hands down sounded the best between all colleagues double blind testing. (and Lavry, Weiss, Apogee, Apollo, and a few more were involved.)

The point to the Apogee statement was this - During that test I involved my old Rosetta and 16X's too along with their new line up (Essemble/Symphony, etc..). First off, their newer line up sound like cardboard across the board. The older units ended up placing above them. The newer Apogees literally ranked bottom of the list in the blind testing. My thoughts are, Apogee fell off, they completely suck now. Sound literally like the recording is on cardboard to me. Hey Prism did too as their newer units (Orpheus, etc.. at that time) sounded good, but no where near as good as the ADA-8XR's.

FYI, The 16X's sounded great. They really do. Just not the best. Apogee now are like kids toys, built for kids. Look like kids marketing, sound even worse. etc.. I nicknamed them Crappy Appy Gee's hence the statement, but I forgot y'all won't get my inside joke ! lol

Originally Posted by psycho_monkey View Post
- why are people sending you stuff with plugin processing applied? Get them to send it raw! Yes - in theory, exporting said crappily processed file as 32b could stop clipping if they’ve overshot. You know what’d be better? No processing at all!
I know I know bro... tell them every time. I swear its what I said, cause most the time it is vocals and I think (just my theory) they are embarrassed about something and they fix it. You know how vocalists are, its like, "no one can hear my voice like this, I only send out my voice fixed up a little".

I even tell them the situation (exactly what I am telling you guys) and say don't worry, it will be higher quality in the end to NOT process anything with your stuff, leave all that to me, I don't judge. Don't worry. But it still happens.

But it still happens. See unfortunately when working for money (not my own stuff), I have to deal with unprofessionals. They have their rule of thumbs, etc.. and not everyone listens. I admit a lot do, but not them all. (Do you ever work on Soundbetter?) Thats where it happens the most.

Originally Posted by mattiasnyc View Post
You are missing my point I think.

My point is that you thought you were right about how bit depth works, and now you seem to think that maybe you misunderstood how that works. Since you just came to that realization you should also consider that perhaps you were also misunderstanding how sample rate 'works'. Because when we sample we deal with bit depth and sample rate, and if you misunderstood 50% of that it's entirely possible that you misunderstood the other half as well.

So it's not about it being "a entirely different subject", because first of all it isn't, and secondly it's all math and technology and if you 'miss' the one you could have missed the other.

I'm actually not entirely opposed to more outlets offering higher resolution audio, I'm just saying that your original post and argument really only boils down to the old tedious debate over sample rates. So any of the million threads about it would do.. if we wanted to talk about it that is..

If all it is is letting us know higher res downloads/streams are now more available then of course we appreciate the info.
Ahhh, yes sir I see your point now and I do that as well. People lose credibility once they are wrong about something. So I see your point.

I want to comfort you though, my statement about sample rates is entirely different than this mystical 32bf thing. I have spent years testing, proving to others within my testing and I also understand it 100% better than the unicorn of 32bf. lol So trust. That one I got.

Also I want to reiterate its not about telling people higher sample rates are better, I just threw that in on the side cause that's my character, lol The real thing/point was/ is like you said, informing us all higher res downloads are offered now and to me that's super cool, because the outlets like iTunes, etc. beforehand weren't even at a 16bit 44.1 level, so the point isn't about that verses higher formats. Its about the crappy mp3 offerings getting better and I think that's a awesome step. And I'd like to spread the word for us all to support it. I'm buying songs now I don't even want just to get the numbers high ! Please everyone do the same is all I am saying. Lets change the world !

About the actual endless meaningless debate of sample rates though - and what this thread is not about nor did I mean it to be, I think I got people riled up over something I did not mean to be a focus - That said I want to say this and be done with that part of the subject. I will admit others might not find important about what I find important (smoothness of delivery timing and plugin reverb tails) about 96k sessions over lower SR sessions, and that's why the debate can be. I get that and I am not trying to change anyones mind. But there can be/should be no debate amongst intelligent people saying that a 96k session doesn't have PROS and cons from 48/44.1k sessions. (everyone just thinks there is the con)

There are pros, that's no opinion, that's fact. Whether your into that difference is a opinion and whether you can even feel/hear that difference I speak of(and no its not high frequencies or any frequency for that matter) is another monkey wrench. I would just like to state that.

Originally Posted by mattiasnyc View Post
As far as I know it is correct that if you go above zero using floating point, and then save that signal in a floating point file, then 'yes', importing that file into a new application that again uses floating point will allow you to avoid clipping.

The recent devices that record to 32-bit float, targeted to location-sound recordists in film/tv production btw, will allow signals 'above zero' for lack of a better word. When importing those files into for example Izotope RX we can see them clipping hard, but by simply lowering the level the dynamics are regained. That's the result of recording to 32-bit float and then importing that into an application that properly supports it.

I suppose the thing to consider here is that the increased dynamic range we get from 32-bit float is really completely unusable at the very end of the process when we actually listen to audio, which in turn means that the much more limited range we need to stick to is.. well.. the limit. In other words we still have to get down to below 0dBFS sooner or later, and going above during production becomes 'questionable' from the standpoint of "best practices". It's really not a good thing to become too used to going above zero is what I'm saying.

But anyway, I think you're partially correct about those implications, if I understood you correctly.
You did understand me correctly. And from what PM said too I know you guys are against it, but it is possible even though I misunderstood 32bf to a extent, that one part of 32bf is indeed helping me in some of my particular workflows and I shall remain to use it then.

Its like a safety net.