RME ADI-2/4 Pro SE - RIAA mode measurements

[badboygolf16v]

@MC_RME often posts here, perhaps he would like to comment. I would imagine he put more than a little bit of thought into the design of this product. I see no reason not to trust him.

 

[sarumbear]

Italic text in blue are copied from: https://www.rme-audio.de/downloads/adi24pro_e.pdf Normal text are my comments.

To be compatible to the varying output voltages of different MM cartridges the ADI-2/4 Pro SE has several selectable gain levels in RIAA mode: +14 dB, +20 dB, +26 dB, +32 dB and +38 dB.

LPs show very different levels of more than 10 dB, from very low (depending on age and genre) to very high (45 maxi singles). MM pickups also very much differ in output voltage. Even third party replacement needles can lead to level deviations of a few dB. In short: a universally usable RIAA preamp has to support a wide level range, and this is exactly what is achieved via the selectable gain of the ADI-2/4 Pro SE.

Sensitivity: +14 dB: 20 mV, +20 dB: 10 mV, +26 dB: 5 mV
Maximum input level @ +26 dB: 28.4 mV, -28.7 dBu
Signal to Noise Ratio (SNR), 20 Hz-20 kHz, +26 dB Gain: 80.5 dB, 85 dBA

The universal level MM amplifiers are test is at 5mV input. That should be the reference level, i.e. the nominal level you will set your system at. It is -18dBFS (US) -20dBFS (Europe). This allows 18/20dB headroom, which correlates to crest factor of 8-10. The 5mV level above has no reference but the levels given for the XLR/TRS inputs are referenced to 0dBFS so I assume it is the same here.

The the input level 5mV at 26dB gain corresponds to reference level of -20dBFS (EU).
They maximum input level at 26dB gain is 28.4mV, which is -5dBFS.
Which is 15dB headroom.

A very affordable device does better than this!

Discrete analog RIAA preamps without adjustable gain need a lot of headroom to avoid distortion at the highest level of a critical record/pickup combination. In addition, the request for a very high headroom is often found, because allegedly crackling and pop noises of the record reach very short, but much higher levels. These would then become more audible in attached devices due to overdriving. That this is not the case becomes obvious when using the ADI-2/4 Pro SE – only on relatively quiet records pops occur higher than the useful signal, and reach only a few dB higher levels. A headroom of more than 6 dB is not necessary, the often rumored 20 dB are a nothing more than a waste of dynamic range.

In other words the entire recording & broadcast industry are wrong and that 20dB headroom they built into every system and standardised on is just “a waste of dynamic range.”

In addition, there is the inevitable reality check: recordings with 6 dB of unused headroom are 6dB quieter than other recordings that (typically) reach 0 dBFS. That may still be acceptable, but with 12 dB or more it just becomes annoying and unusable in practice.

How can a company like RME can place such an illogical text on their user manual beggars belief. Hey guys, have you heard about the thing called a volume control?

The available digital headroom scales with the selected gain, and is visible and can be checked in the level meter at any time, i.e. is adjustable by the user according to personal preference. Due to the unusual concept of the ADI-2/4 Pro SE, the analog headroom is beyond good and evil – based on 5 mV it’s 41.3 dB (580 mV).

No it is not. I have no idea how they calculated the 41.3dB figure but I showed you my calculation above that shows just 15dB headroom.

 

[sarumbear]

@MC_RME often posts here, perhaps he would like to comment. I would imagine he put more than a little bit of thought into the design of this product. I see no reason not to trust him.

I can only comment on the information I read from their material. If I am wrong show me how and I will apologise and thank for teaching me.

 

[badboygolf16v]

I have no idea how they calculated the above figure

Maybe, when you find out, things will become clearer.

 

[sarumbear]

Maybe, when you find out, things will become clearer.

Very funny - not!

If you disagree show me my error.

 

[LTig]

View attachment 249943

I can't believe that this is written by a company who is obviously competent in engineering. They assume the crest factor of music is 0db or just one!

This is not what' s written in the cited text. It's all about peak values.

 

[badboygolf16v]

Very funny - not!

If you disagree show me my error.

I don't know, and neither do you. Until RME explain it's probably better to wait.

 

[sarumbear]

I don't know, and neither do you. Until RME explain it's probably better to wait.

I show my calculations. On what grounds you refute them?

 

[sarumbear]

This is not what' s written in the cited text. It's all about peak values.

Please explain.

 

[LTig]

Please explain.

This is difficult because the cited text is a good explanation. I have no clue how you come to your view that RME thinks that music has a crest factor of 0 dB.

 

[sarumbear]

This is difficult because the cited text is a good explanation. I have no clue how you come to your view that RME thinks that music has a crest factor of 0 dB.

It is not difficult if you know, and no, their explanation is not good nor is correct. Here is what they say:

…recordings with 6 dB of unused headroom are 6 dB quieter than other recordings that (typically) reach 0 dBFS. That may still be acceptable, but with 12 dB or more it just becomes annoying and unusable in practice.

The loudness of a recording has nothing to do with its level reaching 0dBFS, nor any other dBFS value.

Loudness is measured in LUFS. The relation between maximum dBFS and LUFS is called crest factor. A track that hits 0dBFS may sound quieter than another one that doesn’t even reach -10dBFS.

I urge you to read about Crest Factor. Here is a good description by a mastering software publisher.

 

[morillon]

"
Thus, the number alone cannot be used to make a qualitative judgment without knowledge of the material it represents, and in fact, a number is never truly good or bad—it’s just a number!
"
;-)
go in an efficient and factual manner,
somes know lp, music or torture tests, to advise the owners of these new adi to observe whether or not overloads in overmodulated conditions?

 

[LTig]

The loudness of a recording has nothing to do with its level reaching 0dBFS, nor any other dBFS value.

The only word in the cited text hinting to loudness is "quiet". What the text says is that if a recording contains 6 dB of unused headroom (meaning the highest peak is at -6 dBFS) then it sounds quieter compared to a recording whose highest peak is at 0 dBFS. What it forgot to state is that they talk about recordings with similar crest factor. I assumed this to be the case. My Bad .Yet I still don't know how you reach the understanding that RME thinks music has a crest factor of 0 dB.

Hence if you record a piece of music once with 6 dB headroom and once with no headroom and you play them back at the same volume the recording with 6 dB headroom sounds quieter. With 12 dB headroom the difference will be higher - and you waste ADC resolution because you use 2 bits less than the ADC offers. Not that I think that this is a problem when recording vinyl with the RME; you could probably waste 8 bits and amplify digitally by 48 dB and still hear no difference in SQ (which might be just what RME does internally - I did not spot an analog preamp with adjustable gain in front of the ADC in the block diagram posted somewhere above).

 

[sarumbear]

The only word in the cited text hinting to loudness is "quiet". What the text says is that if a recording contains 6 dB of unused headroom (meaning the highest peak is at -6 dBFS) then it sounds quieter compared to a recording whose highest peak is at 0 dBFS. What it forgot to state is that they talk about recordings with similar crest factor. I assumed this to be the case. My Bad .Yet I still don't know how you reach the understanding that RME thinks music has a crest factor of 0 dB.

If a track measuring -6dBFS is 6dB louder than the one that measures 0dBFS the crest factor of the track should be 1 = 0dB. dBFS can only equal dBLUFS (the loudness) if the crest factor is 1.

I did not spot an analog preamp with adjustable gain in front of the ADC in the block diagram posted somewhere above).

From the user manual. They call it a buffer & ASP.

Thanks to MM compatibility, the ADI-2/4 Pro SE does not require an external preamplifier anymore. It has suitable input impedance, RIAA equalization, and the necessary preamplification. p16 RIAA Mode OFF, +14 dB, +20 dB, +26 dB, +32 dB, +38 dB. Default: Off. Activates an RIAA preamplifier and equalizer with selectable gain for turntables with moving magnet cartridge (MM). p23 The RIAA mode is a digital function realized by the ASP (Audio Signal Processor) in the ADC chip p108

 

[LTig]

If a track measuring -6dBFS is 6dB louder than the one that measures 0dBFS the crest factor of the track should be 1 = 0dB. dBFS can only equal dBLUFS (the loudness) if the crest factor is 1.

The text says that the recording reaches 0 dBFS compared to one that reaches -6 dBFS (with 6dB unused headroom).They talk about the peak levels, not the loudness. At least this is my understanding in this context since headroom is the difference between 0 dBFS and the level of the highest peak.

 

[sarumbear]

The text says that the recording reaches 0 dBFS compared to one that reaches -6 dBFS (with 6dB unused headroom).They talk about the peak levels, not the loudness. At least this is my understanding in this context since headroom is the difference between 0 dBFS and the level of the highest peak.

The sentence "recordings with 6 dB of unused headroom are 6dB quieter than other recordings" is about loudness. What can the word "quieter" mean otherwise?

Anyway, I said what I had to say in public. RME is wrong and if they want to argue their case they can do that on this or other thread.

 

[LTig]

Anyway, I said what I had to say in public. RME is wrong and if they want to argue their case they can do that on this or other thread.

Fine with me, let's agree to disagree.

 

[sarumbear]

Fine with me, let's agree to disagree.

You are gentleman, sir. -- I hope the pronoun is correct, its not easy to guess the bear's cloths, not that it means anything these day

 

[badboygolf16v]

I show my calculations. On what grounds you refute them?

You don't understand how the ADI 2/4 works. So how can you argue it doesn't achieve its claimed specs? Options are a) RME are correct b) someone who doesn't know how the 2/4 works is correct. Both are possible but one is more likely.

 

[DrCWO]

Thanks' for the RIAA review. You did a great job

After this review I immediately got an ADI-2/4 to replace my ADI-2 Pro for two reasons:

• It offers a second symmetrical output at the rear so I can connect my Krell Evolution 2250e to drive my horn speakers and also my active Woofers as I run an 2x2 active system. With the ADI-2 I had an additional symmetrizing amplifier SAM-1Bs/2-0 from Funk Tonstudiotechnik in the chain to generate the symmetrical signal for the woofers from the asychroneous headphone output 3/4.
• It has RIAA so I can directly connect my turntable with MM Shure V15IV instead of using the Pro-Ject Phono Box RS as I did before.

So with the ADI-2/4 I have the possibility to operate two devices less!

I combined the ADI-2/4 with the rooExtend-Box and this gives me a complete stereo system. With this I have Phono, CD and iPhone playback plus some nice remote control. All powered by Roon! So XO and FIR Room correction runs in Roon, using the input from ADI-2/4 for phono and an attached Apple Super Driver for direct CD playback using a USB HUB. My wife can even stream from her iPhone via Roon and ists DSPs!

I could not find any sonic improvement with the ADI-2/4 compared to ADI-2 but using the DSP RIAA of the ADI-2/4 improved playback of LPs a lot! In the past I ran input and output of the Phono Box via XLR. Now I plug my turntable directly in the ADI-2/4. But be careful. The 45kOhm input impedance is only true if you connect async from the turntable to ADI-2/4. Getting in via XLR you have 90kOhm. To compensate that I soldered a 90kOhm resistor between Plus and Minus into the XLR connector of then turntable.

Listening to some LPs my wife came down in the living room and asked me what I did that the LPs sound so transparent now! She's not addicted to audio as I am, but she heard it straight away.

So great job RME and thanks' for this nice feature that allows me to shrinks my stereo to the minimum possible. rooExtend-Box, ADI-2/4, Krell Eveoution 2250e plus active woofers. As Roon based this is all you need to get a real good result Maybe I will replace the Krell by two Benchmark AHB2 some day but for today I love my setup and there's not much to complain.

Best DrCWO