Hi @DJ_Big_Blender, if you enable the Mixer section when in the Pro Mode, Classic Mode or 4 Deck Mode, you will be able to see the Auto Gain levels for each channel visually. The system automatically sets this value during the track analysis and you cannot for example change the overall analysis offset.
Within Settings>Sound there is an option to Save and Restore Manually Set Gain Values if you would like to override the autogain levels for specific songs
That’s also an issue/problem in the audio section - there’s too little audio headroom. we should be able to set Output headroom to be -18dB(FS) or -24dB. it will help with what I’ve proposed elsewhere in terms of auto gain and LUFS. it will allow accurate output (master) limiting above -6…-3, -2,-1, -0.5, -0.3, -0.1…(although personally I believe any output above -6dBFS is reckless in digital audio…)
if the headroom is set to -18 and the limiter is set as a brickwall at -3, that’s 15dB of headroom…plenty, often more than enough for most djay sets. setting headroom at -24 and having a limiter at the output for -3 would make for 21db of actual headroom. this may be where we derive auto-gain adjustment values for, devs…
is there anyone at algoriddim who is a member of the Audio Engineering Society, or has a degree in audio production? seems to me this is an issue we keep running up against…nobody knows this stuff there. but I’m also not taking NeuralMix into consideration, etc etc…this is not only functionality/usability, but audio quality at stake, and i know algoriddim cares about that. maybe also a bit of an issue confusing power and pressure calculations with reference to the dB scale and dBFS…but thats not surprising with a team of people writing code…
if Algoriddim has $50, they should google this:
AES41-5-2012-r2022: AES standard for digital audio - Audio-embedded metadata - Part 5: EBU loudness, true-peak, and downmix
Always happy to throw my 2 cents in the wishing well, lol.
LUFS vs the output headroom setting is what the auto gain logic is running up against…from where I sit, it’s not rocket science/brain surgery…
yes Please, I do want more headroom - so if they’re going to work at it, making it user-selectable would be ideal. -18 -15, -21, -24 are all good options…as long as those levels match the LUFS of the songs…see what I’m getting at here? I want the average loudness to match 0 dBVU levels, and the auto gain to shoot for that 0dBVU so that we can do our best to avoid hitting the limiter up near 0dBFS
is that clear enough for engineering? Also, hot diggity, what a fast response time! not that I expect a change to be pushed for tomorrow’s show, but…it’s coming because they’re listening!
Just to be absolutely clear, I want the Output headroom setting to represent the (user selectable) number of dB between “nominal” level and absolute clip/distortion, and the limiter to be user selectable dB below the clip point of the master output.
I believe it will make auto-gain (and likely some neuralMix stuff) work better
Played with it on my phone…looks promising for the main rig…
you’re an invaluable resource to us, I hope you realize @Slak_Jaw - we users (well, I do…) appreciate a direct line to the devs like you’re providing…and please thank whomever had the wisdom to make the position available: support is NOT just solving immediate, functional issues or explaining intricacies, it’s listening to the people who use it for commonly desired features/functions too!!
Alrighty - the headroom option…I’ve tried it and I have further thoughts, @Slak_Jaw
From where I stand, this is the best sounding weekend we’ve had in some time. overdriven speakers and circuitry were completely nonexistent after upgrading. Thank You!!
HOWEVER -
24dB of headroom on channels is too much - it causes the gain to get cranked up to max, and faders (channel and master) to be left full up…So, if possible, in an incoming update, either split the track/channel input headroom to be different from master headroom adjustments, or just apply the headroom setting to the Master output (which seems to be the simplest/easiest way to implement what I’m looking for). 12, 15, 18 (the AES standard), 21 and 24 dB of headroom on the master (and only there) should be plenty plenty of choices and make any djay setup work with any sound system/PA.
Earlier discussions involved the loudness measurement LUFS, for automix purposes, which raised the prospect of headroom; now that we’re getting really close to my ideal, I’d like to suggest that the audio engine coding adhere to AES-18 for headroom on both channels and master, with the option to adjust in 3dB increments on the master for systemic purposes, because the gain adjustment on the channels will allow for plenty of up/down as dictated by the material. (play an old ABBA track that’s basically directly transferred from the master tape right next to something from the height of the Waves L1 wars and you’ll see what I mean)
So thank you for the continued attention to djay end users, management and devs…we’re really making a seriously awesome bit of technology together!
Hi @heysoundude, you’re welcome. Thanks for the positive feedback.
I’ve moved your post over to this topic to help keep the conversation organized. In the future, please don’t post these detailed comments/suggestions in the djay announcement posts because they can get easily lost. Make a new topic or post in a related topic like this one. Thanks!
I have passed you feedback onto the devs as well. Thanks for that as well.
what you write here is totally out of my field of knowledge but i am sure it will lead to better quality for everyone. I totally trust on you knowledge and will look forward to enjoying the benefits of it!
I couldn’t find this original thread/topic when I went looking for it, so thanks again @Slak_Jaw I’ll try to be more precise in the future.
If we can get this all nailed down, AutoGain and AutoMix will be industry-leading (which it probably already is because I think djay itself is…but this will increase the distance between it and the next closest competitor)
I’m not sure if I understand all this, but would like to discuss what I tried for the loudness differences. First of all, is the link posted above correct? Looks like it is about meta data in streams and not the dynamic stuff itself? before I spend the 50 bucks which I would btw.
I did add a compressor that regulates both on peak and on RMS. The peak I did set with normal timing i.e. 7/150 ms, the RMS with a long release of > 1 sec. The idea is to create a kind of post processing auto gain on the RMS to regulate loudness. But actually the peaks are mostly still ‘spoiling’ it, only if there are no peaks the RMS can ‘win’ … And e.g. old school techno or trance that were not remastered recently still sounds dull. … Maybe just a silly attempt …
