I bought an album on Qobuz and this is the spectrogram...

The files are supposed to be 44/16 lossless.

Screen Shot 2022-06-26 at 5.02.15 PM.png

Screen Shot 2022-06-26 at 5.03.30 PM.png

Was I duped?

 

Can’t see the links

 

I see the links, they are identified as 16 bit, 44.1kHz stereo. Seems correct? Interesting, I’ve never seen that display (spectrum vs. time). Cool.

 

Why do you think you may have been duped?

It's difficult to tell anything based on the low resolution of the images and the way Audacity does spectrograms. But at first glance those spectrograms look fine for lossless 16/44.1

Try using Spek which is able to make higher resolution and higher dynamic range spectrograms.

Also, use a site like Imgur to host the images instead of Google Drive. The Google Drive links will not be accessible to most people.

 

Sorry...for a second, they weren't attached properly.

 

How do those spectrograms sound?

 

Okay. I saw a lot of empty space instead of varying colors and I got worried that I bought something that was "upconverted" as the people say. It's happened to me before and it kinda blows.

 

Beautiful colour.

 

I'm glad someone else here sees as deeply into a spectrogram as I do.

 

They look correct for 16/44. But I see some noise above 15kHz which should not be there on first picture. Coud be sort of interference getting into mix. Second picture shows sloppy mixing with track spliced from different sources. But this is what likely came from studio. This is common in non-audiophile records.

 

That's interesting. This particular album was never released on CD and the only places I can buy it lossless are Qobuz and Tidal. Do you think the flaws in the files are the result of poor mixing or poor quality control on the part of Qobuz/the record label? I get weary about buying downloads instead of doing CD rips but with this, I had no other options. I know a lot of downloads nowadays are equal to CD rips because of the way the music industry has changed, but I also know that depending on the release you download and where you download it, you can get varying results.

 

The amount of empty space you see in a spectrogram depends on the dynamic range that the spectrogram is plotting and also the windowing function used and some other factors.

The more dynamic range that the spectrogram plots the more filled in the plot will be. The default for Audacity is to plot 80 dB along with 20 dB of gain. The default for Spek is to plot 120 dB of dynamic range. You can change the defaults for the Audacity spectrogram to plot 120 dB of range along with using the mouse to expand the image so you get a larger image and can see more of the spectrogram.

Here's Spek plotting 120 dB:



Here's Spek plotting 96 dB:


There is more empty space in the 96 dB plot

 

No. Looks like 16-bit, 44.1 to me.

 

It takes practice and experience to learn how to identify lossy source files using a spectrogram. It's tricky. You have to know what to look for and how to achieve that.

Here's a thread with some info about methods and tools to detect if a file may be from a lossy source: How to know if a FLAC is a real FLAC?

I use Windows. I'm not familiar with the tools that work in Macs. There are tools and utilities available for Macs to identify lossy source. Some are mentioned in the above thread.

 

Another trick you can do is to convert the lossless file to a high quality high bitrate AAC and then plot the spectrogram of the AAC file. And compare the AAC spectrogram to the lossless spectrogram. You'll see what lossy encoding does to the spectrogram. If you see a noticeable change and high frequency cutoff in the AAC spectrogram then the source you got from Qobuz is probably true lossless.

 

What album is this? Qobuz just uploads whatever they get from the record label.

 

Album White Out, George Ogilvie | Qobuz: download and streaming in high quality

 

It is very very unlikely that an album recorded and produced and mastered in 2020 would send lossy files to iTunes and other download stores. iTunes and other download stores require that you send them lossless files. And the studios and mastering engineers and labels know better now that lossy in this day and age doesn't cut it.

It did happen occasionally back in the old days of digital files. Some audio engineers and labels really did believe that lossy was 100% audibly transparent. So no harm done if they sent out lossy files to the consumers. Cause nobody could ever notice. The professionals and labels know better now. And the download stores are more strict about getting proper lossless files.

There are still some lossy sourced files that show up on Bandcamp and such even now. It's rare, but can happen. Happens when an album or song is recorded, produced, mixed, and mastered by the band and they didn't know better. Or when a band releases demo recordings that were done back in 1995 on an mp3 recorder. It really shouldn't happen now with a professionally done recording by a competent band.

 

Since this is not a high-res and Tidal also has it, it is possible that these files are MQA encoded with metadata altered to prevent decoding. This would explain noisy few lower bits.

 

That’s what I thought and I’ve found that there can be weird exceptions, which I was hoping this was not.

 

It sounds considerably good with the exception of some rather flat parts here and there. I hate to criticize it too much because artistically, I think it’s quite rich, but I’m happy, not thrilled, with the digital mix.

 

New releases do occasionally have lossy-compressed portions, sometimes on a single track or two out of a larger album, etc. and it absolutely happens to major artists on major labels. There are countless hit songs from over the past several years that had a lossy compressor at some point in their production history.
That said - I really don’t care that much. I wouldn’t let it bother you.
Think about it. If that’s what exists for that release then that’s what exists. That’s what’s part of the album master.
Does it suck? Sure. But at least what you’re getting hasn’t gone through any additional compression or other processing.
In a sense, even if it has lossy compressed material, your file is still a “lossless” or rather bit-perfect/accurate version of the final distribution master as it exists. That’s as high-fidelity as it gets.
Qobuz is a reliable service that spits out exactly what is provided to them by the labels and such. Tidal is a bit more iffy. For the best peace of mind that you’re getting what you really want, I’d stick with Qobuz.


If you complain about buying lossy compressed audio they’ll probably just refund the purchase. So it’s pretty hard to really lose. But based on the spectrals I’m seeing here, it’s not at all clear cut to me that these tracks were lossy compressed anyway.