Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[MrEWhite]

I have a HDCD CD, Green Day's Nimrod. It properly detects it's a HDCD but it doesn't seem to be doing anything to the audio. It does seem to say "Process HDCD" in the Audio Path, but it's obviously still the original, undecoded audio, as when properly decoded, the dynamic range is significantly boosted. Meanwhile, in foobar2000 with the HDCD component or decoding it with dbPoweramp, it seems to be working fine. Any help?

[MrEWhite]

Still doesn't work with the latest update.

[JimH]

Try the instructions here:
https://yabb.jriver.com/interact/index.php/topic,121124.msg837292.html#msg837292

[MrEWhite]

Try the instructions here:
https://yabb.jriver.com/interact/index.php/topic,121124.msg837292.html#msg837292

The HDCD tag is set to 1 and it says it's processing HDCD, but it's really not doing anything with the audio. I can tell because if I play the same file with foobar2000, I properly get the extended dynamic range.

[Matt]

Care to mail me the file?  I'm matt at jriver.  I'll take a look here and see if it shows anything.  Thanks.

[MrEWhite]

Care to mail me the file?  I'm matt at jriver.  I'll take a look here and see if it shows anything.  Thanks.

I went ahead and sent track one of the album straight ripped from the CD, and one decoded with dBpoweramp.

[Matt]

I went ahead and sent track one of the album straight ripped from the CD, and one decoded with dBpoweramp.

Thanks for the files.  I can see that we're changing the track when we decode the HDCD.  It makes a FLAC encode quite a bit bigger.

I'll look at it more on Monday because I'm a little bogged down with something else today.

Thanks again.

[MrEWhite]

I just noticed an HDCD change in the latest version, but it still doesn't properly decode.

[Enger]

I have version 26.0.30 64-bit running on Win-10.
I too see problem with JRiver recognizing HDCD.

I purchased the Audio Fidelity version of the Montrose album.
Foobar2000 analyzer shows all tracks are HDCD.  (none are using peak-extension)
DBpoweramp with the HDCD add-on is able to recognize the tracks as HDCD and transcode them to 24-bit PCM.  (the log file shows each file was recognized as HDCD)

I turned-on the HDCD detection in JRiver.
When I run the audio analyzer on JriverMC against the original (not converted) CD audio files, it does NOT mark the files as HDCD.  (value is left set at 0).

So Foobar2000 and DBpoweramp both recognize the disk as being HDCD.  BUT JriverMC does not.

[MrEWhite]

I have version 26.0.30 64-bit running on Win-10.
I too see problem with JRiver recognizing HDCD.

I purchased the Audio Fidelity version of the Montrose album.
Foobar2000 analyzer shows all tracks are HDCD.  (none are using peak-extension)
DBpoweramp with the HDCD add-on is able to recognize the tracks as HDCD and transcode them to 24-bit PCM.  (the log file shows each file was recognized as HDCD)

I turned-on the HDCD detection in JRiver.
When I run the audio analyzer on JriverMC against the original (not converted) CD audio files, it does NOT mark the files as HDCD.  (value is left set at 0).

So Foobar2000 and DBpoweramp both recognize the disk as being HDCD.  BUT JriverMC does not.

Those AF HDCD discs don't use any HDCD features that make a difference anyways (might use the Transient Filter, which can't be used on software implementations of HDCD), so there shouldn't be any reason to have them marked as one.

[Hendrik]

Media Center only marks them as HDCD when any of the processible filters are actually active, not only when a disc pretends to be HDCD without using an enhancement filter. So its possible that some other software might detect a disc as HDCD, but if it doesn't use any HDCD features, then MC won't.

[jkauff]

Media Center only marks them as HDCD when any of the processible filters are actually active, not only when a disc pretends to be HDCD without using an enhancement filter. So its possible that some other software might detect a disc as HDCD, but if it doesn't use any HDCD features, then MC won't.

That's a good thing, but unfortunately it leaves the user in the dark. If you don't want to change the label on the UI ("HDCD, but no enhancements used."), you could display that information as an on-screen message. This could either appear briefly, or persist while the disc is loaded.

When the disc itself is labeled "HDCD", the user expects MC to recognize it as such.

[MrEWhite]

Thanks for the files.  I can see that we're changing the track when we decode the HDCD.  It makes a FLAC encode quite a bit bigger.

I'll look at it more on Monday because I'm a little bogged down with something else today.

Thanks again.

With the new update, I still don't think it's being decoded properly. The dynamic range of the test track I sent is still the original, non-decoded one (at least by ear, it's significantly louder). It also seems to be clipping at around 160% if I turn on flat line overflows, even though the original, non-decoded one only peaks at around -0.09 dB according to foo_dynamic_range, resulting in volume reductions of around -3 dB if I turn on clip protection.

[dtc]

With the new update, I still don't think it's being decoded properly. The dynamic range of the test track I sent is still the original, non-decoded one (at least by ear, it's significantly louder). It also seems to be clipping at around 160% if I turn on flat line overflows, even though the original, non-decoded one only peaks at around -0.09 dB according to foo_dynamic_range, resulting in volume reductions of around -3 dB if I turn on clip protection.

You could convert your original track to another format with Convert Format and look at the output file. It should have the HDCD parts decoded. You could then run Analyze Audio on it and see how the peak levels and dynamic range tags compare to the original file. You could also use something like Audacity to see if you can spot any obvious clipping.

If foobar can produce a decoded file, you might import that into MC and compare the Analyze Audio tags. dBpoweramp will definitely create a decode file.

The conversion could be as simple as another flac compression or maybe convert to wav.  Convert Format is suppose to honor the HDCD tag now.

[MrEWhite]

You could convert your original track to another format with Convert Format and look at the output file. It should have the HDCD parts decoded. You could then run Analyze Audio on it and see how the peak levels and dynamic range tags compare to the original file. You could also use something like Audacity to see if you can spot any obvious clipping.

If foobar can produce a decoded file, you might import that into MC and compare the Analyze Audio tags. dBpoweramp will definitely create a decode file.

The conversion could be as simple as another flac compression or maybe convert to wav.  Convert Format is suppose to honor the HDCD tag now.

I tried converting it to WAV and it's still a clipping nightmare, something is most definitely wrong with how JRiver decodes. Also, it should be able to play fine without converting anything. Here's an image of how JRiver messes up the decoding, red indicates clipping. I think this is amplifying by 6dB after decoding when it's not supposed to, I got very similar results to this by ticking the +6dB amplification setting in the HDCD DSP in dBpoweramp.

[MrEWhite]

Above

[MrEWhite]

Latest version (26.0.73) still has a ton of clipping when using Peak Extend on this CD (again, applying +6 dB of boost when it's obviously not supposed to I think).

[JimH]

Are you trying to play the CD or a file you've ripped from the CD (and then analyzed)?  I don't think the first will give you HDCD.

[dtc]

Are you trying to play the CD or a file you've ripped from the CD (and then analyzed)?  I don't think the first will give you HDCD.

He is showing decoded HDCD that looks like it has significant problems. Someone should look at his decodes.  From the April 10 post, it looks like Matt has the ripped files.

[Paul S.A. Renaud]

I noted the following. I have a couple of Rosemary Clooney CD's which are HDCD. JRiver's audiopath recognised the CD as HDCD and says Process HDCD when playing but the output is still 16 bit (on JRiver and on DAC). However, I re-ripped the HDCD with DBPoweramp (with DSP HDCD) and JRiver recognised the new version as 24 bit (and no HDCD). However the DAC display shows  44.1 K and constantly switching between 16 and 24 bit. I had never seen anything similar before. From the sound I did not hear any noticeable, so either Process HDCD of 16 bit by JRiver is done well (however the DAC still says 16 bit output after decoding HDCD) or the whole HDCD is a 95% defacto Hoax (like MQA).

[MrEWhite]

Are you trying to play the CD or a file you've ripped from the CD (and then analyzed)?  I don't think the first will give you HDCD.

The album ripped to FLAC then analyzed. Again, I think it's decoding it, then boosting it to +6dB when it's not supposed to.

[Matt]

MrEWhite, could you email me a file where you think the volume is wrong?  We add +6 dB to the filter because that was necessary to make the input and output volumes equal.

But if there's a file where this is a problem, please send it my way (matt at jriver).

Thanks.

[Matt]

Next build:
Changed: Removed the +6 dB shift we were applying to HDCD decoding.

[MrEWhite]

MrEWhite, could you email me a file where you think the volume is wrong?  We add +6 dB to the filter because that was necessary to make the input and output volumes equal.

But if there's a file where this is a problem, please send it my way (matt at jriver).

Thanks.

The Nice Guys Finish Last file I sent you should show should this occurring. If you need it again I'll resend it.

[Matt]

The Nice Guys Finish Last file I sent you should show should this occurring. If you need it again I'll resend it.

I just sent you an email with something to test.  Thanks.

[MrEWhite]

I just sent you an email with something to test.  Thanks.

Your test fixed the issue!