Help with flac files and upsampling

:smiley: I have been racking my brains to find a logical solution to a problem that I hope someone will be able to assist me with with. I’m not a “technical” person so please bear with me :unamused: .
Firstly, I normally keep the default sampling rate as 32bit floating point 44100htz then high quality conversion sample rate as best and dither set to none for processing 16 bit files and Dither also set to none but, will use Triangle for any 24 bit file conversions that I want to convert to 16bit after changing the project rate to 44100khz from say 48000khz

Also my preferred file types are nearly always lossless flac or wav files
Recently, I have found that after doing some “lossless audio checking” with a well known lossless checker programme it returns the status in nearly every case as “Upsampled”. I do quite a lot of remixing or remastering which will involve the manipulation of the original lossless flac file by changing the eq frequency or amplifying the end result to say -1db before exporting the file .
upsampling.jpg
What is perplexing is that in the example shown, I get two different results but have only made a few tweaks to the original source file and save it in various WIP version to play back, listen and then select the best one. It seems to be happening more and yet I cannot understand why. Sometimes I have remastered a whole album and the results come back again as Upsampled. I know that sometime these programmes can report a “false negative” but is there any need to be concerened that I am not getting a truly lossless file and whether it really matters, or am I missing something here that I should be doing in the chain or process from start to finsh. I realise that I have only provided limited information here, so if you require more clarification let me know. It would be interesting to learn if anyone else who does mixing or audio enhancements has experienced the same situation. Thanks in advance

I know that sometime these programmes can report a “false negative”

These programs don’t “know” anything. They can get it wrong and they don’t know what it sounds like. Lots of noise & distortion will probably give you a better report. You can make a lossless white noise or pink noise-file that’s no fun to listen to! A high-definition solo piano recording with very-little high-frequency energy might give you a bad report.

It’s usually easy to fake-out these tools if you try. For example, you can start with an MP3, decompress it, use an “exciter” effect to add high frequencies, and maybe blend-in some white noise and now you’ve got a “high definition” file. Or if you start with a 16-bit file you can up-sample to 24-bits and change the volume very-slightly and that will fill-in the extra 8-bits with what looks-like useful data.

Where are your “originals” coming from? Are you recording from a microphone? Ripping CDs? Recording streaming audio? Something else?

but is there any need to be concerened that I am not getting a truly lossless file and whether it really matters,

If it sounds good it sounds good. If it sounds bad it sounds bad. :wink: I assume your remastered recordings sound better to you?

As long as you keep it “CD quality” or better and you don’t use lossy compression you can assume you aren’t doing any “damage”. (Except you could degrade the sound with effects/processing/editing. :wink: ).

The thing you can’t know is if it could sound “better”. i.e. If you buy an MP3 from Amazon or an AAC from Amazon the only way to know if you’re loosing sound quality to compare it to the CD (preferably in a blind ABX test).

changing the eq frequency or amplifying the end result to say -1db before exporting the file .

Amplifying by -1dB isn’t that useful without knowing where you’re starting from. However Amplifying/Normalizing to -1dB or (0dB) after equalizing (or other processing) and before exporting can prevent clipping (overload distortion) in the exported file.

Thanks for your detailed reply Doug

You asked me:

Where are your “originals” coming from? Are you recording from a microphone? Ripping CDs? Recording streaming audio? Something else?

Essentially,I normally take a lossless flac file and run the Audio checker first to see if it is “Clean”. I then go about trying to improve the sound quality and add EQ by using a VST add-in equaliser I have three different ones that I can choose that I load from within Audacity.

That usually reduces the volume of the file in most cases, after it has carried out the audio changes I like for the song, and hence the reason why I then boost the amplification to around -1db (after the file has been EQ’'d). I do this because I rip cds and make my own cd compilations. Is that still ok or should I set it to “0” db instead ? As far as the sound quality goes, I get excellent results and do post my work to various sites. I might add that I keep the original flac file in it’s previous state, so that I can make several different files for comparison to hear and then select which one I prefer the best.

What still perplexes me the most, is that I often check what other people have posted when I download their stuff and 9 time out of 10 they all appear CLEAN, so that is why I am at a loss to understand how theirs is CLEAN and not mine. I know that might sound a bit “obsessive” :unamused: but that’s the whole reason that I first became aware of this anomaly. All that I can tell you is that some of the people I have spoken to use a MAC instead of Windows for their processing. In fact, I sent one of my files to a friend who has a MAC and he did a minor tweak and it came back “CLEAN” but could not tell me why?? I’m wondering whether I should alter the floating point in Preferences from 32bit back to 16bit and try that?

Thanks