... by a country mile.

a good recording at 44.1khz will out do a slightly less good recording at a higher rate by a country mile!

Hang on... there's an echo in here.

Hopefully it sounds alright at 44.1kHz...

THIS!

I produced an album for a major label at 88.2kHz.
Everything REALLY sounded better during the process, especially EQ-plugins, time-stretching. So, yes indeed a difference.
However as the standard format still is 16bit 44.1 for CD and Streaming services, the final result after printing it down for final delivery was like, ok was it really worth the hassle...half the power of everything...no it was not.
There is a point if the result should be delivered at 88.2 or higher but for Spotify etc, No way. I am back at 44.1.
I am using the AVID HD I/O for all tracking. Still as good as it gets.

Interesting, i myself came to the exact opposite conclusion (not saying you are wrong though, all up for discussion):

My PC has all the power to run my sessions at any rate so that´s no issue to me.
Since processing works better at higher rate i use that.
When downsampling for CD i still get the benefit of mixing at higher rate as this final step will not alter the sound that much.

In other words: Going from high to low ONCE in the final step when no more processing will happen is better than doing all the editing, processing and mixing in the lower resolution to begin with.

I do classical music and jazz recording, so usually i am going for a very natural sound and high dynamic range. That might make a real life difference compared to rock, pop or rap, where coloration is a stilistic element and there is even less right or wrong.


If i find the time i might want to make a little experiment:
Record a session at 96 - Mix - Print - Downsample
Same session at 96 - downsample each track to 44.1 - load the same mix as above - Print
And then compare and look at the difference by flipping phase.
Question: I think when doing this in reaper i only need to change the rate of the converter since reaper will know that i´m running 96khz files in a 44.1khz session - and will downsample on the fly. Anyone know if that´s the case?

And one more thing:
Of course 44.1 sounds amazing all by itself on modern converters!
We have truly come a long way!

Did you A/B the 24/88.2 with the 16/44.1 release?

Did you A/B a 24/88.2 -> 16/44.1 conversion with something that was 16/44.1 all the way through?

Or was it just the idea that you had to reduce it to 16/44.1 that got you down?

Tape sounds better than vinyl. So should we just cut to vinyl or should we keep the better quality orignal for a future releases?

If the 24/ 88.2 sounds so much better, then the artist/band/managment is really missing a trick by not releasing it in any form.. There are loads of outlets for hi-res!

Is dithering still a thing when going to 16 bit?

I remember an article interviewing the guy from Meridian (the hifi company and digital pioneers - the same guy who came up with MQA) in the 90s that talked about getting a theoretical effective 19 bit resolution playback (in certain psycholacoustically important FR ranges) from a 16 bit CD via dithering techniques alone (from a 24 bit source and when played back on a 20 + bit DAC). Something to do with maths and noise.. I could be confusing the figures with HDCD, but still it was something similar.

TLDR at the bottom.

I'm that one weird kid who did some recording with a tascam 688 and a Guillemont maxi studio isis. I've used the 688 as just a mixer for 0 lag and mainly recorded into the isis . It uses Bur-Brown 20-bit converters and records at 48k on all 8 inputs at once. The unit sounds surprisingly very clean and very good. My guess is due to those converters and the breakout box. Only issue is the unit only works on windows 98/me. Plus, I have a billion cables to hook it all up.

I haven't used it in a long time and upgraded to a an old 2626 interface and a usb daw controller, which is really kind of a pain. Since then I barely have the time to record much of anything with it.

I'm finally out of poverty and am looking to do some recording and maybe get the old band back together. I'm sick of cables and patchbays and plastic daw controllers and just want something that works and sounds good and I really prefer having a real mixer. I know, from the isis, 44.1 and 48 sound good. And I know, from research, most newer preamps and converters sound very very good these days.

So I'm upgrading again and selling all the old equipment, maybe send some to storage. I've done the research and I'm awaiting the arrival of a Behringer x32 Producer that should replace nearly everything and it's, you guessed it, 44.1/48k. Cost me less than $1200 from Thomann.


TLDR: I'm replacing all my stuff with a Behringer x32 producer. It's highest sampling rate is 24bit/48k.

He was talking about noise shaping which was a new thing in the '90s. Dither is still necessary to have minimum distortion.

For me this whole topic is pretty important because...... I use the same computer since 10 years. I thought about buying a new one many times but every time I did I got to the conclusion that I dont need one. For my purposes, my computer is absolutely fine......BUT I record at 44.1k/24bit.

If I want to switch to 88.2k or 96k (my interface supports this), my computer is not sufficient anymore, just too slow..... what to do? Can someone decide for me please....

untill recently i was using an i7 2600k, so thats about as old as your machine.
I had no problem with 96khz.

This depends on your sessions of course.
You should try raising your buffer. With 96k you can double your buffer and still have the same latency.
If (like me) you monitor through your converters direct monitoring or through a mixing desk, than there is no need for super super fast latency.

To play amp sims or synths live you need that very low latency. Guess what, my ryzen 3700x is not much better in that regard compared to the old sandy bridge i7.
It has a lot more compute power, but the realtime performance is quite similar...

Also your system might be bottlenecked somewhere else... but that´s only guessing.

If your plugins alias, yes it will make a difference.

All those production/mixing moves you made at 88.2 (sounding better) will have a positive cumulative effect on the sound.

Having a track produced and mixed in higher quality and then SRC to 44.1 is a different prospect to producing and mixing at 44.1 (i.e. plugins sounding worse) from the get go. Especially if you use a top SRC tool like Weiss Saracon or SoX.

Your efforts were not in vain.

I have done extensive ABX double blind testing and I cannot hear the difference between 44.1 and 96.

Therefore I mix and record 44.1/24 - my Crane Song converters do a fine job at this sample rate.

Not to mention most of my favourite albums that I still use a reference recordings were recorded on Sony DASH recorders at 16 bit 44.1 !!

Of course what makes them great is the stunning song writing, performing, stunning studio spaces and brilliant engineers and producers. I keep that in mind.

Equally, my inability to hear a difference may be a function of being 57 years young :-)

I have an i7-2600 (16GB RAM) and am pretty much at the limit with 44.1khz/24bit already. Mainly because of the numerous amp sims. I know, I can record guitars with direct monitoring, without virtual amps. But I had to learn that it is not only more funny to record via amp sims. It does affect my playing so I prefer recording via an amp sim to make the music as good as possible. When the project weights more than 8-9GB, I start to deactivate audio tracks I dont need anymore in my DAW. Otherwise it gets VERY tedious to work with the project. As soon as some limit is reached, everything becomes very slow in Cubase. I guess its normal. Then, I deactivate some tracks I dont need anymore and the project runs at normal speed again. With this workflow, everything works fine at 44.1khz/24bit. I use SATA3-SSDs so maybe I could try to record at 88.2khz once but I am not very optimistic on this to be honest. The amps, I guess, will need a lot more cpu power.

My best sounding productions have always been 44.1. Go figure. I mostly use plugins with very little to no aliasing though. All by design

I used to record at 44.1k mixes were good BUT when I went to 88.2 my mixes became great. Just easier to mix plugins sounded better and reverb sounded like reverb. I would go 96k but I don't see the need as I'm getting great resolution now. Now I know everything is going back down to 44k but why not capture as much information as you can from the jump. I don't think I'll ever go Beck to 44.1 as I definitely hear a difference. Is it significant? NO but it's there. I could mix at 44.1 and lose no sleep I made great mixes at that in the past.

We are kind of hijacking this thread, but to reply:
There are a number of things that come into play when you want realtime responce for a lot of ampsims and synths.
Like i said, i did not have the impression that going from an i7 2600k to an ryzen 3700x made a huge diufference in that regard.
By the time i am getting dropouts and have to dial back the audio buffer the cpu is only utilized around 30%, and usually running around 3ghz - faar from it´s potential.
The reason is that my system has other bottlenecks when it comes to realtime performance.
The OS is one factor. The video and lan driver another.
A major difference can be experienced between different audio interfaces.
A PCIe card will perform a lot faster than USB2.
Some drivers a simply better than others.
The DAW is a factor. Reaper would propably allow for more tracks at 88.2khz than cubase. It´s definitly the case with reaper vs live (i don´t know about cubase).

Things to try in order to increse your realtime performance:
- maybe test a different amp simulation. could be whatever you are using just doesn´t have good performance
- freeze tracks so you only have one amp sim "live" at one time.
- disable unused channels on your interface
- disable LAN and then turn off any antivirus software and windows realtime protection. Before going back online be sure to activate protection.
- kill all unnecessary startup processes - although that might not be a big thing any more. Not sure.
- if it´s really getting laggy you can hide the mixer and any plugin gui to regain some power

But first of all try for yourself if your ampsim really sounds (much) better at higher resolution.

It´s not necessary to spend big cash and nothing is holding you back from producing great music!
Maybe the highs will sound ever so slightly different. who cares in rock music...
It has been said before: A lot of folks are happy with 44.1khz
At the same time there is a lot of voodoo and missunderstandings going around.
Clockspeeds is at least a thing you can measure (well how about that!) and uncover the difference with an analyser or a nulling-test. Just saying...

Thx yes I will finish producing the actual EP in 44.1khz/24bit and try to go to 96khz for the next EP. If it does not work I probably buy a new computer. I think this is not too off-topic and if it is....this is somehow what a web board is about imho. We do not discuss the possible re-election of Trump here

What I will NOT do is setting any amp sim plugins oder virtual instruments to "low quality" or similar. If I do this, I bet, one day I will find out that I sent some guitar track or virtual instrument exports to the sound engineer with the feature "low quality" on. Nooo way but thank you I will try the other hints