Sources do sound different, scientifically proven by measurements. You just have to do the right measurements and/or look at them correctly but they do sound different.
The issue is, that people only record test sounds and (usually) only test at 100% Volume. I test gear by playing music and/or white noise and at different volumes.
I play the music/noise, record the outpt and use a Software called "DeltaWave" to check the difference (Delta) between the original music and the output from the source gear. So i know exactly what changed in the soundwave.
I do this in two ways. I record directly the output from the source and in a second test, i record the output with different receivers (9 Ω IEM, 32 Ω Headphone, 470 Ω Headphone)
That way i found out why my Walkman sounded different than my DAC/Amp and different to my DAC/Dongle (all 3 sounded different) and why exactly it sounds different.
I went to different sound science guys and nobody believed me, they said its impossible, one of the devices is broken and so on. But i did know they sound different, i could distinguish them in a blind test so i bought measuring gear myself and tested with that.
And the measurement did show it. Conclusion, nobody tested any of these devices in real-world conditions. Most testers just connect and let the Software do the rest and wait for the result. Turns out if you use the devices how you use them in the real world, they perform different than in those synthetic test cases.
What annoyed me was, that my 50$ Dongle performed best and close to perfection. Really really good. The 2000$ DAC/Amp performed worst and the 3500$ Walkman was somewhere in the middle.
During further testing, i also found out why. Most devices i tested using this method perform worst at low volume. The louder you make them, the better (objective, not subjective, so closer to the original thing) they sound. At 100% Volume for example, the DAC/Amp performed very good, pretty much perfect, but then i tested at real world listening volumes.
It had a completely different frequency response on -90db, -70db (where IEM are already very loud), -50db (where most Headphones are already very loud), -30db and 0db
Its almost perfect at 0db but sounds broken at -90db with a very strong roll-off in bass and lower mids and a boost in the upper mids and that did show up in the measurements.
After further investigation, it turned out that the volume control of this DAC/Amp was the reason for this. And its like that by design.
On the Walkman it turned out that Sony did not implement an output snubber which caused the signal to overshoot and so change the frequency response.
In these tests by just playing music, the best devices (objectively) have been those with the lowest power output, as you can use them at relatively high volumes and digital volume control used with 24/32bit up sampling (The free upsampling that comes with your OS, no special device!) where they perform closest to ideal.
So in my personal experience the louder you can make something, the better it will sound. The more power something it has, the less loud you can use it (unless you like to listen at 160db while blowing out your eardrum and the IEM/Headphone at the same time) which will lead to distortion and or change the frequency response due to analog volume control or other factors.
So sources do sound different, it can be measured and, in my experience, (objectively!) the less power something has and the simpler it is built, the better it sounds. The higher the output power and the more fancy the components, the more will the soundwaves differ from what is originaly in the song.
If you subjectively prefer that, is on a completely different piece of paper. But if you want objective best performance, in my personal experience:
- Digital Volume Control
- Upsampling to 24/32bit (the free upsampling your OS does is sufficient) to prevent rounding errors of the digital volume control
- Output power that allows you to use the device at high volume with your receiver without destroying your hearing
But even then, you have no guarantee. Without measuring in real world circumstances, you simply don't know and maybe you don't care. Maybe you like the sound with higher distortion, why not? But sources can sound different.
That said, i also tested a lot of sources so far that all performed 99,99999% identical. So sources can sound different, quite significantly, but at the same time, most sources are pretty much identical.
Example:
DAWN PRO <--> Walkman --> Small difference
DAWN PRO <--> TA-ZH1ES --> Big difference
DAWN PRO <--> M11 Plus LTD --> Small difference
DAWN PRO <--> M11 Plus ESS --> Identical
DAWN PRO <--> MIAD01 --> Identical
DAWN PRO <--> Tanchjim Stargate II --> Identical
DAWN PRO <--> SMSL SP200 --> Small difference
DAWN PRO <--> SMSL SP400 --> Identical