Loopy Pro: Create music, your way.
What is Loopy Pro? — Loopy Pro is a powerful, flexible, and intuitive live looper, sampler, clip launcher and DAW for iPhone and iPad. At its core, it allows you to record and layer sounds in real-time to create complex musical arrangements. But it doesn’t stop there—Loopy Pro offers advanced tools to customize your workflow, build dynamic performance setups, and create a seamless connection between instruments, effects, and external gear.
Use it for live looping, sequencing, arranging, mixing, and much more. Whether you're a live performer, a producer, or just experimenting with sound, Loopy Pro helps you take control of your creative process.
Download on the App StoreLoopy Pro is your all-in-one musical toolkit. Try it for free today.
Are IOS Apps designed to be used @ 48khz Resolution only?
Was wondering if anyone know if all IOS apps are designed to be used at 48 khz only or 44.1 khz? What about higher resolutions like 96 khz?
I've noticed some apps like Nembrini effects act weird or sound weird at the higher resolutions...any thoughts?
Comments
They shouldn't be, and it's rare that they are these days. However, some host / app combinations have buggy behavior coordinating sample rates. I've never seen that with Nembrini apps though. That's odd. Which host?
Hi Wim, yeah it's weird. I was using thr Delay 3000 and rack delay and for some reason they don't sound as warm as when I play them at 44.1 or 48k. But when I use iFX Rack Delay, it sounds warm sort of Speak, in other words full, when I use them at the lower or higher resolutions. I'm going to continue to test it out with different configurations. I love the Nembrini Rack Delay though, it's got that analog warmth that I'm after so hopefully it's just something on my side.
I don't see how sample rate can have anything to do with audible difference in "warmth". I thought you must be referring to crackling and distortion. But I believe you.
Could be that you're hearing anti-aliasing. You must have far better ears than I do. 😉
Lol! Distortion would be cool. Might be anti-aliasing but who knows. I've been working a lot on getting the best tone out of my setup. But I wonder if maybe I should just stick with 48khz. I thought the higher the resolution the better for tone but I'll keep checking it out. Thanks Wim!
Yes, 44.1 or 48khz will tax your CPU much lower.
Regarding „sound“ today almost everything considered „critical“ is oversampled by the app/plugin and essentially all output converters also oversample to defeat aliasing.
It‘s a huge difference from the early days of digital processing.
(fun fact: such artifacts are highly sought or even re-progrrammed)
You may have run into a situation of double anti-aliasing:
the program is designed to oversample, but you feed it a signal that already has a high sampling rate. This may result in strange side effects, if this particular case isn‘t anticipated (and handled) by the software.
It‘s an individual case specific to each software, not a general rule.
iOS famous for running audio apps at incorrect pitch at 96khz sample rate, mainly due to it not being picked up in beta testing as mostly people use 44.1/48khz. can be fixed by the dev though.
Running at 96 kHz is much more taxing on the cpu than 44.1 or 48k…and the audio files are twice as large because you are processing twice as much data.
Unless you are achieving noticeably better audio quality at 96k, I’d run at 44 or 48k
Interesting
I'm going to switch to 48khz and see if I notice a quality difference! I know that a lot of the high end guitar effects modelers run at 96khz so I thought maybe that's the reason they sound so good. But to me, as long as the sound is spot on as far as amplifiers is concerned, them the rest is icing on the cake. The effects I really want to shine the most are my Pitch Shifter, delays and Reverbs. So far the one that totally blows me away is the Nembrini Klon Clone. It really makes any amp I place after it to really shine!
Too bad I can't test this whole weekend as we'll be so busy at a marriage conference 🙃
But next week, man, I'm all in.
The main reason is probably that doubling the sample rate will have the side effect of half the latency
yep, strymon reverb pedal are all 96khz, personally I find modeled instruments like pianoteq sound better at 96 khz.
So if my phone's cpu can handle the cpu load at 96 khz do you recommend using that instead? I always wondered if the 24 bit was better than 16 bit but now it's the resolution that perhaps is most important to sound quality?
I don't think there's any double-blind tests that even show conclusive evidence that 48 kHz sounds better than 44.1 kHz, let alone 96 kHz vs 48. So essentially, all you're doing is doubling CPU and battery usage for no perceptible benefit.
BTW: both the sampling rate and the bit depth are resolutions. The sampling rate is the resolution in time, while the bit depth is resolution in amplitude / voltage.
The bit depth has a very strange unit for historical reasons -- "bits". It's strange because for every 1 point increase (say, 16 bits to 17 bits), you get a doubling of resolution. While for the sampling rate, it's linear (96 kHz = twice the resolution of 48 kHz).
Dropping out of Teacher Mode again 😄
trust your ears in all cases for music. if you think it sounds better that's the way.
That applies to most of current digital processing, but there are exceptions.
Which are so obvious that it doesn‘t need double blind testing.
A synth with the original Waldorf wavetables (8-bit and grainy by character) is a tremendous source of aliasing at 48khz.
This effect is fully removed by doubling the sample rate to 96khz, grit is almost turned into silk.
But now you don‘t hear the character of your orginal synth anymore... even the filter has less to chew on.
From an „objective“ pov 96khz improved the sound... with a significant loss, though.
If it‘s considered loss or gain is an entirely subjective matter. Let your ears decide.
A similar difference exists for the PPG synths, but in this case W. Palm used state of the art modern „strategies“ for his software products, while his old hardware features the unavoidable grainy character of digital processing at the time of design.
Which one is better ?
Of course I was referring to "normal" situations where the algorithms are designed so that the sampling rate is not "intentionally part of the sound" (I hope I expressed this properly 😄). i.e. situations without "hacks". (not meant as criticism...)
Yes, such cases have become to rare that I had to go back to that old Waldorf thing...
all non-linear processes (like filtering or saturation/ drive) where non-harmonic distortion can be problem, will always highly benefit from higher sample rate .. as a solution is often used oversampling …
When you decrease amount of non-harmonic distrotrion, result sound may really sound “more warm” .. so ses - higher processing frequency (or oveesamoling, or both) can lead to more “analog warm” sound …
try in obsidian create some patch using FM oscillator with higher operator ratios and higeher modulation amounts, also some feedbacky so iz starts to touch “noisy” fm sounds area - rhen record few notes on higher octaves - then render it as 44khz and 96khz - the difference may be absolutely major.
filtering, overdrive, oscillators - all those thing have substantial quality difference when samoling rate doubles ..
Oh wow! Such great information.
Yeah, I'll definitely see if I hear any difference. Since my iPhone is processing analog overdrive pedals coming in from the Xtone Pro audi/midi interface, I figured once all that gets mixed with the digital world of iPhone apps, I wanted the highest possible resolution.
But because I heard some delay apps and even reverb sounding a somewhat metallic instead of the warmth I've been used to hearing, I figured the resolution could be a culprit. Or perhaps the apps themselves were not being programmed to effective work at higher resolutions. But I can't wait to dig in this coming week and do some real hands on tests.
I would have someone play the recordings back to you without you knowing what you are hearing. Non-blind tests often yield results heavily influenced by expectation.
On a pro audio discussion board several years ago, there was a lot of discussion of the difference in the noise floor of two recordings…one made with AC power and one with battery.
A week after the samples were posted and discussed, it turned out they were the same sample file which had accidentally been uploaded twice. Of the 10 or so pros commenting , only one observed no difference.
This is a side effect of the control design of (most) reverb/delays‘ feedback path, intended to suppress frequencies in the high mid range, which are responsible for this impression.
It‘s a fairly complex/tricky process to „do this right“ and sometimes even a business secret.
But it‘s essentially independent of bit depth and sample rate.
OK, but (note, I'm not a DSP expert!) this rather seems like a limitation in Obsidian to me, i.e. it should internally use the "appropriate" amount of oversampling automatically, no matter the output sampling rate. I don't think ANY digital sound generator should sound noticeably different at different output sampling rates (as long as they're at least around 40 kHz). i.e. Obsidian shows "implementation details" to the outside world, which IMO is not correct. (it's an amazing synth overall, but this I think is actually undesirable).
Yeah, but they shouldn't! If there's a significant difference in sound between the internal processing at 48 kHz vs 96 kHz, then it should always use at least 96 kHz internally.
BTW, there was a very similar issue in NS1, but regarding buffer size. Many things were processed exactly once per buffer, so that when you set different buffer sizes, things like LFOs and even automation sounded noticeably different. Again, I think this shouldn't be the case. NS1 could have always used the minimum possible buffer size internally (say, 64 samples) for its processing and then concatenate the buffers to the "actual" buffer size.
But then I must admit all of this is REALLY PICKY for someone like me who hasn't ever implemented a complex DSP algorithm from scratch 😄 so... SORRY!
@SevenSystems
many synths are not using oversampling even on desktop and they all always benefit from increased host sample rate .. I used example of NS2 just to illustrate why using higher internal sample rate makes sense, cause it is easy to test there, not arguing if it is good or bad that NSdoesn’t use oversampling (except of overdrive FX where s optional switch for oversampling) .. of course if synth / fx uses oversampling, difference between 44/49 and 96 is not that significant .. but many don’t …
Of course to do entire internal processing at 44/48 sample rate and then just upsample result mixdown to 96khz (i believe this is doing most of iOS daws which are allowing 96khz mixdown) is complete nonsense with zero benefit.
Obviously higher sample rate, same as oversampling, comes with price - increased CPU usage. So it's debatable if it should be automatic standard everywhere. At least on mobile devices before M1 era it is always good to have it just optional, at least in half (if not more) use cases it really doesn't matter ..
Yes, that's all true of course... I think the OP was refering mostly to reverbs and delays, which probably are much less sampling-rate-dependent than the examples you mentioned... they mostly just copy raw audio data around
well, they do some filtering as well sure, but I doubt the difference will be huge here...
I'm not sure, but I'm wondering if what @Tones4Christ is really saying is they're sending guitar signal through their audio interface at 96kHz attempting to get higher quality, but that some apps are sounding less warm, perhaps because of downsampling from the higher rate to a lower internal rate.
If so, It seems to me that the simple answer would be just to not do that. I doubt an electric guitar signal,\ is going to be noticeably enriched for having been grabbed at a 96kHz.
Said with absolutely no technical knowledge of the subject at all. 😂
Well when it comes to sampling electro/accoustic instruments, there is noticeble diffrence between 44khz and 96khz sample rate (especially in upper sprecrum range) ... I would say when sampling external instruments, do it always when possible 24bit / 96khz ...
Of course when audio processing is then made in 44 (or 48) - which is the case on iOS - lot of benefits are lost, but you still have material in good quality which you can use in good quality audio enviroment
i mean if you love sampling, sampling hardware and then pitching/mangling samples then 96khz is the way or even 192khz.
We're talking about an electric guitar signal processed live through distortion and amp simulators, reverb, delay, etc.
I did some more tweaking and was listening to other worship guitar players going thru their setup and hearing their tone. Of course, a lot of them had effects pedals that each cost around $400 or more each!!!
So I did my tweaking at 48khz, and at 96khz, and believe it or not, I couldn't tell the difference in warmth too much. But there definitely was more sharpness in the tone, hard to explain but just enough to have me feel more happy at my tone at 96khz.
But the big difference was that at 96khz, vs 48khz at 256 buffer, the CPU was hovering around 47%!!!
I always thought the higher resolution would choke my cpu but it didn't. I'm using an iPhone 12 pro max. I was so blessed by my mother my past birthday and got a refurbished iPhone to use on my pedalboard.
Oh, and one of my cables was messed up! That cable was creating some kind of phase issue and this threw me off!!! I removed it and the tone came alive!!! Even Stratosphere reverb...oh my goodness! I love it! This Reverb saved me $400!!! The Strymon Bigsky Reverb pedal is so expensive!!! All I need was that Cloud Reverb! And Quantum Delay, by
@gsdsp totally blew me away! It literally saved me another $400 from using the Strymon timeline delay pedal. All I wanted was the Tape delay! I am using two instances infront my ToneX dual amp iPhone setup!!!
So grateful for these amazing and awesome developers!!
If you kept the buffer the same at 96kHz then you also halved your latency. 😎