New Fade Effects (poll)

Which fade effects would you like to be included with Audacity?

0 voters

I would like to propose replacing the current “Cross Fade In” and “Cross Fade Out” with two other effects,
There are many fade effects to choose from, but given that the Effect menu is rather long already I doubt that increasing the total number of effects would be acceptable.
This poll therefore allows 2 votes per person.

It would be helpful if, as well as voting, you could say briefly why you have voted for the options that you have.
If you would prefer some other option (not listed on the poll) please say what your preference would be in your reply.

Briefly …

Pro Fade Out
I have been using this for a fair while and it quickly became my fade-out of choice for working with music - I am keen that other users should easily be able to share in this easy to use effect that produces such excellent results.

Adjustable Fade v30
Though I will probably not need this myself I recognize the need and the semand for an adjustable fade. I have been testing variou versions of this effect and v30 has an interface that (even) I can readily understand - plus it looks “spiffy”

The problem with this is that these fades do not easily replace the existing cross-fades - or does AF v30 aim to fill that gap? What about your easy-to-use Crooss-Fade-Classic? I suspect that without “something” being labelled “cross-fade” this will be seen as a regression to be avoided.

Peter.

An important consideration here is that “Cross Fade In” and “Cross Fade Out” do not perform crossfades.

Cross Fade In/Out provide one, fixed contour non-linear fade shape that may sometimes be suitable when crossfading.
My estimate would be that this fade profile will be appropriate in less than 10% of cases.

Adjustable Fade (version 30) provides a variable contour fade shape. Whatever your needs may be for fading in, out or crossfading, the adjustable fade is likely to be able to provide what you need.
My estimate would be that this can satisfy over 90% of cases.

I don’t think that we need to provide a “crossfade” effect in order to replace Cross Fade In / Out, because they are not crossfade effects.

Do you see it as a regression?

Personally no - but then I never have a use for those “cross-fades”

Peter

I’ve voted for “another version of Adjustable Fade” (i.e. with more concern shown for fade types) and for Pro Fade Out.

I would “personally” rather see a “Musical Fade In” than “Pro Fade Out” for these reasons.

  • I think Pro Fade Out still sounds “mechanical” rather than “acoustic”. I had never seriously tried it before now, but I would want another + 4 dB at least at the end to get “my type of fade out”. The tail is too long and quiet for me.
  • I think it would probably be easier to find a “musical” fade out elsewhere than a musical fade in (In other words, a musical fade in could be plugging a gap in the market).

Nonetheless I actually voted for Pro Fade Out because I recognise my views are oddball and Pro Fade Out will probably be more used than a musical fade in. I continue to argue for a musical fade in to pair with pro fade out.

I think Text Envelope may now be hard to argue for as a shipped effect, which disappoints me somewhat, unless we subsequently decide to “extend” Adjustable Fade with multiple envelope points. Perhaps we should instead be arguing for an accessible (and more flexible) C++ interface for Envelope Tool?

We do have to be careful about removing Cross Fade In and Out, not because they are very useful for cross fading but because they are a long established feature that for all we know people may be using and liking. My implied vote to replace them requires that something similar does replace them.


Gale


PS Can whoever else voted for “another adjustable fade” please say what they want?

Please listen to these two samples.
Would you say that the fade out of one is noticeably “better” than the other?

(personally I think both sound a bit abrupt)

I don’t like either - you’re right they both sound a bit abrupt.

Peter.

but do they sound very different? Do they both sounds about the same? Does one sound better/worse than the other?

Why is that?

One of the things that I like about “version 30” is that it is virtually a “half way house” between a “simple fade” and a full “text envelope”. However, before spending a lot more time on a text envelope effect I’d really like to get a lot of user feedback from the wider Audacity public about the “Adjustable Fade” effect. This has been sitting on the burner for far too long. Let’s get it out there and see what comes back.

Because you have repeatedly said that multiple envelope points would overcomplicate Adjustable Fade. I am not totally sure I agree, but I gave in to that, partly because it would be more useful IMO to improve accessibility in Envelope Tool. Whether we gave it an accessible GUI or not, it would be a huge improvement to be able to see the actual exact level of each envelope points. See Missing features - Audacity Support .

There are always two views as to whether to “get it right first time” or to release an incremental version that we know won’t satisfy everyone.

I suspect with this effect, feedback may be hard to come by.

I’m sorry if it was not clear in the vote that I support #30 except for the reduced range of fade types, but as #30 stands I could not vote for it “as is”.

My vote makes clear I want an Adjustable Fade shipped. Like you, I’ve spent way too much time on it for it not to be shipped.


Gale

To add multiple envelope points in addition to the current features would (obviously) make the effect more complex. However, do we really need all of the features of an “Adjustable Fade” in a “Text Envelope” effect? I think not. For example, I don’t see any need for a “Mid Fade Cut/Boost” control in a “Text Envelope” plug-in (though I do think that it is essential in an “Adjustable Fade” effect).

I don’t see a “Text Envelope” Nyquist plug-in as a viable replacement for “Adjustable Fade”, but as a complementary effect. It may be possible for a C++ implementation to provide the functionality of both effects, but I doubt that it can be done satisfactorily in a Nyquist plug-in (unless the Nyquist plug-in interface is developed further).

But I expect even harder to get feedback if nothing is released.

Can you can describe precisely what features you need?

No I think not too which is why I dropped that suggestion.

And given how hard it also is to get released stuff improved and re-released, we should try to get this as good as possible IMO.


Gale

I completely agree that we should never release shoddy work, but I think this plug-in is far from shoddy.
It should not be difficult to re-release improved versions. Sometimes the fear of regression gets in the way of progress and development - imho we sometimes need to be a bit more bold.

In practice it is clearly easier to add features than to remove features because there is less risk of the change being seen as regression.

If at any point a C++ implementation to replace Adjustable Fade/Text Envelope appears (which I hope will eventually happen), then as long as it is at least as good as whatever is shipped at that time, I will fully support “upgrading” to the C++ version because a C++ version has the possibility of a much better user experience. It should not be made difficult to do that.

Can you can describe precisely what features or changes you think version 30 needs to be “as good as possible”?

I am not calling it “shoddy” - you know none of your work is that - but I think a user frustrated by his/her ignorance possibly might call it some names if it is not “de-geeked” a little. I think there are now two clear choices of how to “de-geek” and I hope one of them will be OK for you ( Adjustable Fade - #213 by Gale_Andrews ).

As you know re-release can be difficult. Delay and Vocoder need re-releasing to give two examples but will probably have to wait a long time.

The quality standard for the 2.x series explicitly tolerates only “trivial” regressions:

I have to be “hot” about that.

It’s kind of difficult if the “regression” might make things worse for 20% but better for 80% of folks. The 20% should not be more than minimally inconvenienced by the change, and it should be discoverable for them to adapt.


Gale

B is terrible, hardly any fade out until “chopped off” at the end.

A is less bad, quite good until “chopped off” at the end.

What lies behind the question?



Gale

One of them is “Cross Fade Out” and the other is Adjustable Fade with a high Mid Fade Boost setting (+90 if I recall correctly).

The point is that they are very similar, but neither of them is very good (for this audio)

They both end too abruptly which is the most obvious thing wrong and with the Adjustable Fade effect that is easy to remedy - just reduce the Mid Fade Boost a bit.
High Mid Fade Cut/Boost = more “abrupt” at the end
Low Mid Fade Cot/Boost - more “gradual” at the end

Delay was updated recently - it’s still not perfect but until someone writes a better pitch shift algorithm for Nyquist or codes it in C++ it is probably as good as it can be.
Vocal Remover, Silence Finder and Sound Finder are my next targets.

See my response on the Nyquist thread: Adjustable Fade - #214 by waxcylinder

Peter

Are you going to put me out of my misery and tell me which one was Cross Fade Out? :wink:

The fade seemed to be unusually long. Was that part of the problem?

Maybe a “preset” would work well too if there are sufficient of them? :ugeek:


Gale

That was originally my intention - but I’ve forgotten :blush: :stuck_out_tongue:
I think that the Cross Fade was the least bad of the two.
Adjustable Fade with a lower Mid Fade setting sounded a lot better than either.


For the type of fade used, the selection was not ideal.

I deliberately did not “optimise” anything - as a new user might do, I just did a quick “click, drag, apply”, then used the same selection for both. You can usually get away with such a slapdash approach when using “Pro Fade Out” because of the gentle taper at the end, but with a distinct (abrupt) end the selection needs to be a lot more precise and considered.

That little “blip” of the next note is not pretty. With these fade curves, just undoing and trimming the track just a little shorter to avoid that “blip” would have made a big improvement.