You are not logged in. Please login or register.
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → Bitperfect test and harmonic distortions
Is it possible to pass bitperfect test despite of very high level of 3rd harmonic distortions for -1.1 dBFS and above?
Some users say that streamer X passes RME bitperfect test. Some others report very high signal distortions with their devices.
The bit test validates “bit perfect” transmission of “digital audio data” and runs in the “digital domain”.
Transmission of zeroes and ones from the audio player to the ADI-2 Pro shortly before D/A conversion.
In the digital domain are no harmonic distortions, this is only a data transfer of zeroes and ones.
Comparable to opening a PDF document from a USB stick.
PDF document is a binary combination of data which represents some objects like text or graphics. There is the same analogy for digital audio. I didn't mean analog distortions, I meant a binary combination of data which represents harmonic distortions in the source audio stream.
If you decode a sweep signal in its binary form on the streamer and then capture a binary form being a result of any processing made on the streamer, you can compare both in the proper software and for example measure the level of distortions introduced. That is what I was asking about: is it possible to pass RME test when significant amount of 3rd harmonic distortions has been introduced during decoding of the test file?
> is it possible to pass RME test when significant amount of 3rd harmonic distortions has been introduced
> during decoding of the test file?
What should introduce harmonic distortion here?
What do you mean with decoding?
The test files for the Bittest are wave files, this is a digital uncompressed audio (PCM encoded).
The purpose of the test is to validate that nothing alters digital audio data during the transport from the player to the ADI-2.
If there were something that adds something to it (be it harmonics or whatever)
or in other words, changes the digital data (be it in the player or in the audio subsystem of windows etc)
then the Bittest would fail.
> is it possible to pass RME test when significant amount of 3rd harmonic distortions has been introduced
> during decoding of the test file?What should introduce harmonic distortion here?
What do you mean with decoding?
By decoding I mean process of converting the audio stream being performed by a streamer to the raw PCM form which is later delivered to the DAC. Digital audio data can be untouched or altered during that process.
I observe the situation where my device adds mentioned distortions and on the other hand some users report that their devices pass RME test. I could imagine that my steamer is faulty, but there is at least one user who has got similar measurements on his device.
As I do not know how exactly RME bitperfect test works apart of information that specific bit pattern is being searched constantly in the incoming PCM data, I'm curious if it's possible to pass the test with harmonic distortions introduced on the player side
If there were something that adds something to it (be it harmonics or whatever)
or in other words, changes the digital data (be it in the player or in the audio subsystem of windows etc)
then the Bittest would fail.
Are you absolutely sure that adding, removing or changing any data decoded from wav test file will not allow to pass the test?
There is a bit pattern repeated 20 times in the wav. Will the test fail if silence area between those patterns is altered any way?
... is it possible to pass RME test when significant amount of 3rd harmonic distortions has been introduced during decoding of the test file?
No.
Any type of signal alteration in the digital domain gives a fail in RME BitTest.
Only a completely unchanged, original signal gives a “BitTest passed“ result on ADI-2’s screen.
To say it the other way round:
If you see “BitTest passed“ you have the best signal quality possible and can exclude all detrimental effects like distortion.
Any type of signal alteration in the digital domain gives a fail in RME BitTest.
Only a completely unchanged, original signal gives a “BitTest passed“ result on ADI-2’s screen.
Is it effective to the whole content of a wav test file which must be unaltered for a pass result?
Or just one of 20 bit patterns inside must be unaltered to show "passed" message?
If I, for example, shorten the test file to leave less bit patterns in it, would it pass the test?
KaiS wrote:Any type of signal alteration in the digital domain gives a fail in RME BitTest.
Only a completely unchanged, original signal gives a “BitTest passed“ result on ADI-2’s screen.
Is is effective to the whole content of a wav test file which must be unaltered for a pass result?
Or just one of 20 bit patterns inside must be unaltered to show "passed" message?
I never analyzed the BitTest WAV in such detail, don’t even see a practical reason to do so.
If you want to know, test it.
I bet a changed BitTest WAV will not pass.
What I can say:
If you want to catch random issues, repeat the BitTest multiple times, but insert a break of about 4 seconds between repeats, or the results screen will not show up correctly, according to developer “MC” Matthias Carstens.
I never analyzed the BitTest WAV in such detail, don’t even see a practical reason to do so.
If you want to know, test it.
I bet a changed BitTest WAV will not pass.
Unfortunately I cannot test it as I no longer own RME ADI-2. That's why I'm asking instead of testing it myself.
What I can say:
If you want to catch random issues, repeat the BitTest multiple times, but insert a break of about 4 seconds between repeats, or the results screen will not show up correctly, according to developer “MC” Mattias Carstens.
I guess you've meant looped wav playback without breaks.
Does that mean "passed" screen will not show in case of passed test? Or "passed" screen will show for failed test?
BitTests passed message will not show for every repeat.
But if you don’t own an ADI-2, what are you heading for in such detail?
BitTests passed message will not show for every repeat.
But if you don’t own an ADI-2, what are you heading for in such detail?
As I've tried to explain at the beginning some users face distortions generated by the streamer which can be measured and they look exactly the same. On the other hand some users say that their devices with the same FW can pass RME bitperfect test.
I'm trying to figure out how it's possible.
...some users face distortions generated by the streamer which can be measured and they look exactly the same. On the other hand some users say that their devices with the same FW can pass RME bitperfect test.
I'm trying to figure out how it's possible.
It’s not possible on the same machine, but you seem to talk about different users with different machines.
RME BitTest WAV is a full scale signal, so if there is distortion BitTest will fail.
There are a lot of other reasons why distortions might appear, without knowing further details it’s impossible to examine this from the far.
E.g. if you use a Yamaha receiver, these usually can’t take more than 2.5 Volts input in the analog domain.
Easy to get distortions with too high levels on those.
It’s not possible on the same machine, but you seem to talk about different users with different machines.
RME BitTest WAV is a full scale signal, so if there is distortion BitTest will fail.
There are a lot of other reasons why distortions might appear, without knowing further details it’s impossible to examine this from the far.
E.g. if you use a Yamaha receiver, these usually can’t take more than 2.5 Volts input in the analog domain.
Easy to get distortions with too high levels on those.
Yes, different users and the same model of a streamer with probably the same FW. All those distortions are in the digital domain.
If it's not possible to get the test signal distorted and pass the test same time, we will have to look for an answer somewhere else. I'm glad to know that RME test is useful in such scenarios.
Thank you for your help.
But, as said at the beginning, those distortions are only for high level signal, no less than -1.1 dBFS. RME test signal seems to be on much lower level.
But, as said at the beginning, those distortions are only for high level signal, no less than -1.1 dBFS. RME test signal seems to be on much lower level.
No, I just checked, RME BitTest WAVs are digital fullscale.
They cover this case.
No, I just checked, RME BitTest WAVs are digital fullscale.
They cover this case.
Are you sure about it? REW tells me it goes only to 45.3 dB for 96/24 test file.
KaiS wrote:No, I just checked, RME BitTest WAVs are digital fullscale.
They cover this case.
Are you sure about it? REW tells me it goes only to 45.3 dB for 96/24 test file.
-43.5 dB RMS probably, but 0 dBFS Peak.
As I said, yes, sure, I just checked in reality, RME BitTest WAVs are indeed digital fullscale, no question.
It looks like you are wrong, unfortunately. One of the users has measured digital output of the affected device and he had got heavy clipping distortions for standard 1kHz sine wave at -0.2dBFS. The same unit was able to pass RME bitperfect test.
It looks like you are wrong, unfortunately. One of the users has measured digital output of the affected device and he had got heavy clipping distortions for standard 1kHz sine wave at -0.2dBFS. The same unit was able to pass RME bitperfect test.
If you had the ADI-2 you could look by yourself.
Who knows whatever the mentioned user did, BitPerfect WAV is in fact digital fullscale, 0 dBFS.
To see is as easy as a look at the meters, that I did, and did again right now.
While passing BitTest, ADI-2’s Meter PreFX shows OVR with BitTest WAV, clear evidence it’s 0dB FS.
If there is clipping in the signal path to ADI-2 BitPerfect Test cannot pass.
If the clipping happens later in the signal path, e.g. due to a DSP effect, BitPerfect Test does not “know” about that.
The problem with your inquiry: you’re taking a position of a man in the middle, so we can’t find out what’s going on at someone else’s place.
If you had the ADI-2 you could look by yourself.
Who knows whatever the mentioned user did, BitPerfect WAV is in fact digital fullscale, 0 dBFS.
To see is as easy as a look at the meters, that I did, and did again right now.
While passing BitTest, ADI-2’s Meter PreFX shows OVR with BitTest WAV, clear evidence it’s 0dB FS.If there is clipping in the signal path to ADI-2 BitPerfect Test cannot pass.
If the clipping happens later in the signal path, e.g. due to a DSP effect, BitPerfect Test does not “know” about that.The problem with your inquiry: you’re taking a position of a man in the middle, so we can’t find out what’s going on at someone else’s place.
We are talking about signal being distorted by a streamer as he has captured the audio on the cloned toslink input passed to usb output. I was capturing the signal of my device similar way and with similar result on my miniDSP Flex. He has provided the captured signal:
https://drive.google.com/file/d/1lLAGeP … share_link
https://drive.google.com/file/d/1Xzvpq2 … share_link
There are no reasons to me to suspect him cheating as he looks to be a kind of very satisfied RME user. Anyway I will try to convince him to post here himself.
I never analyzed the BitTest WAV in such detail, don’t even see a practical reason to do so.
A few details here: https://forum.rme-audio.de/viewtopic.ph … 60#p140060
Bittest is a short click sequence of 100ms length.
The display shows the result for 2 seconds, during this time no further checks are being performed.
KaiS wrote:BitTests passed message will not show for every repeat.
But if you don’t own an ADI-2, what are you heading for in such detail?
As I've tried to explain at the beginning some users face distortions generated by the streamer which can be measured and they look exactly the same. On the other hand some users say that their devices with the same FW can pass RME bitperfect test.
I'm trying to figure out how it's possible.
What steamers? What files from which steaming service? You provide no specifics, data or clarifications. This forum is not for trolling or BS hypothetical irrelevant issues.
What steamers? What files from which steaming service? You provide no specifics, data or clarifications. This forum is not for trolling or BS hypothetical irrelevant issues.
Streamer model doesn't matter. Test file details have been specified few posts above - "1kHz sine wave at -0.2dBFS". Captured response file has been provided as well.
I do not consider asking for answers here as a kind of trolling. It's also neither hypothetical or irrelevant.
yuhasz01 wrote:What steamers? What files from which steaming service? You provide no specifics, data or clarifications. This forum is not for trolling or BS hypothetical irrelevant issues.
Streamer model doesn't matter. Test file details have been specified few posts above - "1kHz sine wave at -0.2dBFS". Captured response file has been provided as well.
I do not consider asking for answers here as a kind of trolling. It's also neither hypothetical or irrelevant.
Again with all the subjective opinions . Other forums on web welcome these undocumented and unneeded speculations .
onlyoneme wrote:yuhasz01 wrote:What steamers? What files from which steaming service? You provide no specifics, data or clarifications. This forum is not for trolling or BS hypothetical irrelevant issues.
Streamer model doesn't matter. Test file details have been specified few posts above - "1kHz sine wave at -0.2dBFS". Captured response file has been provided as well.
I do not consider asking for answers here as a kind of trolling. It's also neither hypothetical or irrelevant.Again with all the subjective opinions . Other forums on web welcome these undocumented and unneeded speculations .
May I kindly ask you to reduce your presence in this thread if you are not interested in any investigation related to the case I had been described? Thank you in advance.
According manual (Pro manual, page 94) ADI has three (16, 24 and 32-bit) bit perfect check circuits, constantly running i.e. waiting their Right Pattern. They are programmed to react these and only these RME´s bittest.wav -test file bit patterns. Their bit pattern is 400 samples long, very specific line of 0´s and 1´s. Like 6400, 9600 or 12800 characters long password.
Check circuits won´t trigger if a bit pattern fed in deviates even by one bit of what they are expecting. Which would be a case if feeding streamer distorts i.e. fucks its output bitstream even by one bit. "Wrong password" i.e no pass in such a case.
Is it possible to pass bitperfect test despite of very high level of 3rd harmonic distortions for -1.1 dBFS and above?
Some users say that streamer X passes RME bitperfect test. Some others report very high signal distortions with their devices.
No.
If this distortion really origins from digital signal, some PROCESSING in this Streamer X must do this alteration i.e. add that distortion there. Some bug in handling of sample rates, would be my quess. Wouldn´t be a mirracle, such a bug-sack peace of shit streamers are on market.
Anyway in that case, when putting bittest.wav on playback and fed to ADI, bittest.wav´s bit pattern would get severely modified due to said bug-processing of Streamer X: ADI´s check circuits wouldn´t trigger i.e. detect it from incoming bitstream.
Yes... there are many kind of Digital Audio Einsteins etc untechnical-morons playing expert-influencers in internet these days... in reality some of them barely qualified to use a screwdriver. 
The picture seems to be cleaner after additional tests.
All the "real" audio files get distortion if their peak level is high enough - at least -1.1 dBFS.
RME test files are being unaltered for whatever reason. Maybe they are too short, maybe bittest is not complex enough, I don't know. That's why they pass the test.
yuhasz01 wrote:onlyoneme wrote:Streamer model doesn't matter. Test file details have been specified few posts above - "1kHz sine wave at -0.2dBFS". Captured response file has been provided as well.
I do not consider asking for answers here as a kind of trolling. It's also neither hypothetical or irrelevant.Again with all the subjective opinions . Other forums on web welcome these undocumented and unneeded speculations .
May I kindly ask you to reduce your presence in this thread if you are not interested in any investigation related to the case I had been described? Thank you in advance.
May I kindly suggest you read the technical replies here( like one next) and go in a different direction—empirical data and technology— to pursue your “issue”
KaiS wrote:No, I just checked, RME BitTest WAVs are digital fullscale.
They cover this case.
Are you sure about it? REW tells me it goes only to 45.3 dB for 96/24 test file.
Kais is correct.
The 24/96 bit file test has the following levels as per Izotope RX:
Peak Level = 0.00dB
Max RMS = -29.01dB
Total RMS = -34.95dB
Kais is correct.
The 24/96 bit file test has the following levels as per Izotope RX:
Peak Level = 0.00dB
Max RMS = -29.01dB
Total RMS = -34.95dB
I know, I've observed that myself as well.
And I've also performed some tests using RME test file as I get ADI-2. I've added fade effects to the 96/24 file, random noise and attenuations. Only 2 of 20 bittest patterns remained unaffected and it was enough to get the message about test passed.
Regarding the issue I started with. That's the fact that streamer was adding clipping distortions to any content played with peaks above -1.2 dBFS, because of DRC limiter implemented. And the device was passing RME test because its bit test patterns were unaltered.
... that streamer was adding clipping distortions to any content played with peaks above -1.2 dBFS, because of DRC limiter implemented. And the device was passing RME test because its bit test patterns were unaltered.
A simple and likely explanation:
The Limiter’s attack time constant is too slow to affect the very short peaks of RME BitTest signal.
It passes before the Limiter reacts.
As you found out, only one pattern needs to pass to make BitTest say OK.
In contrast, a sine wave or music signal has longer duration peaks, triggering the Limiter.
Why not just disable the Limiter/DRC in the streamer?
A DRC with a Threshold Level of -1.2 dB doesn’t make sense anyway.
Why not just disable the Limiter/DRC in the streamer?
A DRC with a Threshold Level of -1.2 dB doesn’t make sense anyway.
It was a FW bug, partially resolved by the manufacturer already. I think it has been fixed due to stubbornness of some users who made proper measurements to show the issue. All of that against some users who were absolutely sure that passing RME bit-perfect test guarantees that such thing couldn't happen.
Well, re-reading the first posts everyone was right. You clearly asked about high order third harmonics added. You did not talk about a complex FX with attack and release times working on the digital signal - and most probably nobody expected this to exist in a streamer device. At least not me.
Well, re-reading the first posts everyone was right. You clearly asked about high order third harmonics added. You did not talk about a complex FX with attack and release times working on the digital signal - and most probably nobody expected this to exist in a streamer device. At least not me.
I had simply no idea what the reason was. All I knew was the distortion effect. We had been told about DRC and a limiter recently by the manufacturer after his investigation.
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → Bitperfect test and harmonic distortions
Powered by PunBB, supported by Informer Technologies, Inc.