Should digital headroom for DACs and DSPs be tested as part of review measurements?

[KSTR]

But ISOs are possibly inaudible.

That statement makes no sense. ISOs are a property of the source material and a fact of life, and the problem is how faithfully a DAC can reproduce them or at least not butcher up the conversion with strong artifacts.

 

[VientoB]

If this problem exists in any musical material, it's not its worst problem. With a probability approaching 1, this material is compressed to the limit and is difficult to listen to precisely for this reason.

Compression is a different thing entirely. Any music which goes up to close to 0db can be affected. Highly dynamic music likely has larger peaks.

 

[Mike123]

Compression is a different thing entirely. Any music which goes up to close to 0db can be affected. Highly dynamic music likely has larger peaks.

Mathematically, you're right. In practice, in highly dynamic music with an RMS of -15 to -20 dB, frequent peaks are impossible by definition. The greater the difference between the RMS and the track's peak, the less frequent and shorter the peaks. Therefore, their negative impact, even if it exists, is much less pronounced. In low-dynamic music with an RMS of -4 to -6 dB, 0 dB peaks occur practically every second. But even they, as I wrote, aren't the main problem in this case and likely won't cause any significant issues.

 

[etc6849]

Part 2 video with audio clips:

 

[MatrixS2000]

dBTP levels above 0dBFS are not compatible with proper mastering practices IMO, unlike what John Siau appears to claim.

Yes, problematic tracks exist. No, they're not the majority.

But so what if a track has 1000 samples above 0dBFS? A typical track consists of around 10M samples. So 0.01% of samples would be clipped.
That's a total of 20 milliseconds of clipped audio scattered across an entire track, or this long (in sum):

View attachment 556151

Most of all though: Where are the ABX tests proving audibility of ISOs in musical material?

Yes, if I had the choice then of course I'd like my DAC to also handle this edge case gracefully.

But is it ThE BiGGeSt PrObLeM in DiGiTaL AuDiO??! Not even close, at least not until we've solved the loudness war.

This - particularly the last line.

ISOs in relation to all samples:

 

[solderdude]

Below a video by Ian Shepherd with ISP explained with examples.
It is a mastering issue and does happen and can be audible.

Basically ... the solution is very simple .... reduce the volume digitally by 3dB before it goes to the DAC or to an ASRC or converting to lossy codecs.
Easy for playback (and a good idea anyway) but reckon no one does this before upsampling or converting to lossy codecs and then the -3dB trick does not work anymore.

I second to add the test as proposed by @KSTR simply to identify DACs that do so (and reckon many do).
In the end it is a mastering issue should be handled in the studio but alas is not always done ... who knew ... and could become audible.

The suggested extra test will at least 'mark' the worst offenders.
Sure... extra work for @amirm and he would be the only tester who tests for this but this will also be the extra value for looking at ASR tests.

Personally I always use digital attenuation anyway and am never bothered by ISP anyway, they don't matter to me nor distract me from enjoying music.

 

[MaxwellsEq]

The suggested extra test will at least 'mark' the worst offenders.
Sure... extra work for @amirm and he would be the only tester who tests for this but this will also be the extra value for looking at ASR tests

@NTTY does some tests for Inter-Sample Overs.

 

[etc6849]

Yes, it is true that it's a small fraction of the album. However, I still want the most accurate analog waveform. Who wouldn't?

The video is very interesting from a technical standpoint too, and the testing John is doing is 100% scientific from an electrical engineering stand point.

You can hear the difference in signals, and you can see the clips on his Rohde and Schwarz Oscilloscope (yellow trace - ISOs). He also says in the video he did this to train his ear, and then test whether it can be heard in their listening room. Will you spot all 1000 instances in a given track, obviously not. Can you be trained to pick out a few on a given track? It sounds like it according to the video...

This - particularly the last line.

ISOs in relation to all samples:
View attachment 560422

 

[solderdude]

Yes, simply lower the volume digitally before it reaches the DAC by 3dB and you have the exact waveform.
Just 3dB lower in level...
It's exactly what John Siau does in his DAC when enabled so can you and anyone else?
John even says this in his video.
Just lower the volume digitally by 3dB and you are safe. To get the same volume as 'normal' operation would give simply turn the volume up 3dB AFTER the DAC.

And ... if you want to up-sample a song, for some reason, or want to use severe lossy compression for some reason:
Lower the volume 3dB first, then upsample/encode, then export and its fine.

Problem solved ... nothing to see anymore.

The actual problem lies with incompetence of producers that make products with ISP over 0dBFS.

Sure... it would be fine if DACs were tested for this and could faithfully reproduce +3dBFS but every DAC already has that but at 0dBFS.
It is the responsibility of the one making the music files (producers) to keep in mind that they should ALWAYS remain below 0dBFS even with calculated ISO (ISP over 0dBFS).
There are plenty plugins that can help/prevent that.
Now it is just a matter of educating producers HOW to use them.

 

[danadam]

You can hear the difference in signals,

I'd rather say "you can hear the signals' difference". And you can, because then it's not masked by the signal itself.

I'm also not sure if I buy his explanation about the distortion spread in time and its audibility. You can see that the distortion before and after clippings looks like a sine with more than two periods per division and one division is 50 µs. That gives more than 40 kHz signal.

 

[etc6849]

Because I don't need to? I have DAC3 B's, plus I stayed at a Holiday Inn last night...

That said, I think it's definitely worth knowing what your particular DAC is doing internally, just so you know whether to lower the upstream digital volume in the first place. Since most manufacturers do not tell you, that's why it really should be tested in reviews.

After all, why would you want to lower volume digitally if you don't need to? Many may have target curves and EQ up stream of their DAC(s). Digital headroom is valuable to them...

Yes, simply lower the volume digitally before it reaches the DAC by 3dB and you have the exact waveform.
Just 3dB lower in level...
It's exactly what John Siau does in his DAC when enabled so can you and anyone else?

 

[popej]

We already have similar thread:

Let's develop an ASR inter-sample test procedure for DACs!

Hi there, It's been a while since I thought about this, but why aren't DACs tested for their ability to handle inter-sample overs (or ISP, inter-sample peaks) in ASR reviews? Here are some established facts: ISPs are very real, and almost any material that's been released after the 90's can...

www.audiosciencereview.com

I use an AVR, which has a DSP before DAC. Most probably DAC headroom becomes irrevalent for this configuration, just like bit perfect playback or resampling. Still I would like to know, which DAC chips handle ISO correctly.

 

[solderdude]

Because I don't need to? I have DAC3 B's, plus I stayed at a Holiday Inn last night...

That said, I think it's definitely worth knowing what your particular DAC is doing internally, just so you know whether to lower the upstream digital volume in the first place. Since most manufacturers do not tell you, that's why it really should be tested in reviews.

After all, why would you want to lower volume digitally if you don't need to? Many may have target curves and EQ up stream of their DAC(s). Digital headroom is valuable to them...

Yep, that gives piece of mind and ensures that mistakes made during the producing stage don't have a possible negative impact on the sound.
The audibility of ISO remains to be seen. Having some insurance this can't be the case is 'pleasant knowledge'.

People not owning these rather expensive DACs can feel confident this won't happen even with their cheap DACs when they simply dial the digital volume down at no cost

Even if Amir would start testing with the 'simple test' in future reviews there would still be questions as to how many dB ISO's are allowed, how audible this is one still would not know for all the previous measured DACs.

 

[etc6849]

I think that is very fair observation, but I don’t think the ~40 kHz frequency of the visible ringing by itself establishes that the time spreading is inaudible.

The orange trace is the difference signal between the clipped and unclipped DAC outputs. The error from the very short clipping event is no longer confined to only the instant the clipping occurred, due to the reconstruction filters in the DAC. The filtering redistributes the error both before and after the clipping event. Note for others: that does not imply anything crazy like the DAC is psychic ; remember, a linear-phase FIR filter introduces sufficient latency to produce this type of pre-ringing... So yeah, the clipping is clearly spread out in time (both before and after).

I'd contend you have a valid point. However, just because the waveform appears to be ringing at ~40 kHz isn't proof that the entire error signal is spectrally confined to 40 kHz and above. The orange trace is a complex error signal of a transient, it is definitely not a stationary 40kHz sine wave. A short, amplitude-varying burst can certainly have a broader spectrum. An FFT of the error signal would tell us considerably more than counting time periods on the scope.

That said, I don't think an FFT of the entire event would completely settle the question of audibility either. It would tell us whether the error signal contains audible-band energy, but not when that energy occurs relative to the clipping event. To really test his explanation about distortion being spread in time and its audibility, I think we'd want to see the spectrum versus time, such as a spectrogram, to determine how much audible-band energy exists before and after the clipping event.

I'd rather say "you can hear the signals' difference". And you can, because then it's not masked by the signal itself.

I'm also not sure if I buy his explanation about the distortion spread in time and its audibility. You can see that the distortion before and after clippings looks like a sine with more than two periods per division and one division is 50 µs. That gives more than 40 kHz signal.

 

[kemmler3D]

That's exactly what I'm proposing. A quick and feasible check to weed out the crap.

But this is the real issue here - if there were real crap out there in quantities worth caring about someone would have noticed. Name a DAC that qualifies as crap in this sense?

Someone raise your hand if you're genuinely bothered more by the sound of overs than the idea of them?

 

[solderdude]

Note that the difference trace is 10x as high (20dB amplification) as the signal traces.

 

[popej]

Note that the difference trace is 10x as high (20dB amplification) as the signal traces.

Like 24dB SINAD.

 

[solderdude]

During an extremely short time period though.

 

[popej]

There are positive ABX results of hearing ISO problems in the thread I have indicated earlier:

Let's develop an ASR inter-sample test procedure for DACs!

Hi there, It's been a while since I thought about this, but why aren't DACs tested for their ability to handle inter-sample overs (or ISP, inter-sample peaks) in ASR reviews? Here are some established facts: ISPs are very real, and almost any material that's been released after the 90's can...

www.audiosciencereview.com

 

[DanielT]

There are two issues here:

1. How big of a deal this is.

2. What is the fix?

In all of my experience playing heaven knows how many pieces of music, I have never, ever heard such an artifacts. I have once, verified that my son had a repeatable case that triggered this on one DAC, and not others. At least I think that was the case.

Now to #2. The only fix available is to assume such clipping exists AT ALL TIMES, and therefore, steal dynamic range from everything played. That is, reduce the max level of the incoming digital signal, and then increase gain in analog domain to compensate, which adds constant noise.

But let's say you want to do that. As John says, media players and streams already have volume control performed in digital domain where you can dial in attenuation. And if you are going to do EQ, then you will have to do that anyway. Indeed, this is why Roon player has dynamic range compensation as a checkbox:

See how I have enabled "headroom management?" That is a constant attenuation which you can dial to any value, include the recommended 3 dB for intersample overs.

Further, if you have just about any DAC that doesn't say otherwise, with a volume control, that volume control is in digital domain and prior to actual conversion, sample rate conversion, etc. To with, here is the ESS DAC:

This is what I have on my desk and since I never play at 0 dBFS, I am covered. And what I recommended people to buy, a DAC with volume control so you don't need a pre-amp either.

If one wanted a "proper" fix, it would require a tool that scans every track and dials in a headroom. Media players could do this easily if this were a real problem. Hardware DACs cannot do this since they don't have memory to store entire tracks, analyze them in advance, and then play.

The only time you may want such a thing would be if you still operated a CD player, fed it to a DAC with no volume control, and then a pre-amp. Since hardly any of us do this, I am not about to force this down the throat of every DAC to reduce its digital output.

So let's not keep bring this up. Benchmark and a few others who do this should have a switch that defeats it and with it, gain another 3 dB of SINAD.

Aha, so that’s how headroom management works with Roon.

With the setup I’m currently using—WiiM’s RoomFit—I determine my volume settings (or management, so to speak) based on @staticV3 test results regarding RoomFit’s EQ headroom found here:

Guide for Managing Volume and Headroom on the WiiM Mini (Measurements)

This is a small excursion into volume and EQ headroom on the WiiM Mini (and by extension, likely other WiiM products). WiiM Mini Firmware: 4.6.805929 WiiM Home Version: 3.5.1 For all tests, I used a -1dBFS Measurement Sweep, exported from REW and played back directly via the WiiM Home App...

www.audiosciencereview.com

Furthermore, since we’re discussing clipping, RoomFit also offers a "Non-Boost Mode" option:
This setting ensures that the sum of all correction values does not exceed 0 dB, preventing distortion and clipping caused by excessive boosting.

Introducing RoomFit™ Settings

TABLE OF CONTENTS Overview Target Curve Frequency Range Max Gain Min Gain Max Q Non-Boost Mode Built-in Mic Compensation External Mic Calibration Subwoofer Calibration Moving Mic Measurement (Beta) Smoothing Tip Overview WiiM...

faq.wiimhome.com